Anticipating the Climate Black Swan

by Judith Curry

I just received the reviews on the manuscript I submitted to special issue of the journal Climatic Change, entitled “Reasoning About Climate Uncertainty.”  Overall, the review was pretty mild.  One comment from the editor was about my paragraph that mentioned “black swans.”

The specific text from my paper:

The framework associated with setting a CO2 stabilization target focuses research and analysis on using expert judgment to identify a most likely value of sensitivity/ warming and narrowing the range of expected values, rather than fully exploring the uncertainty and the possibility for black swans (Taleb 2007) and dragon kings (Sornette 2009). The concept of imaginable surprise was discussed in the Moss-Schneider uncertainty guidance documentation, but consideration of such possibilities seems largely to have been ignored by the AR4 report. A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries. 

The editor’s comment, linked to the last sentence:

Good point, you should elaborate how this would help science, policy

What is a black swan?

The concept of “black swan” comes from Nassim Nicholas Taleb, author of the book “The Blacks Swan: The Impact of the Highly Improbable,” .

The term “black swan” comes from the mistaken assumption that all swans are white. In this context a “black swan” is a metaphor for something that cannot exist. Black swans were discovered in Australia in the 18th century, thereby proving false the assumption that all swans are white. In Taleb’s context “black swans” are rare events beyond the realm of normal exprectation.

From the Wikipedia:

The Black Swan Theory or Theory of Black Swan Events is a metaphor that encapsulates the concept that The event is a surprise (to the observer) and has a major impact. After the fact, the event is rationalized by hindsight.

The theory was developed by Nassim Nicholas Taleb to explain:

  1. The disproportionate role of high-impact, hard to predict, and rare events that are beyond the realm of normal expectations in history, science, finance and technology
  2. The non-computability of the probability of the consequential rare events using scientific methods (owing to the very nature of small probabilities)
  3. The psychological biases that make people individually and collectively blind to uncertainty and unaware of the massive role of the rare event in historical affairs

Unlike the earlier philosophical “black swan problem“, the “Black Swan Theory” (capitalized) refers only to unexpected events of large magnitude and consequence and their dominant role in history. Such events, considered extreme outliers, collectively play vastly larger roles than regular occurrences.

Examples of black swans in Taleb’s context include 9/11, financial collapse, power grid failure, the internet, rapid climate change.

A key element of “black swans” is that while they are unanticipated, in hindsight they events are rationalized to seem obvious (it could have been anticipated with better use of available information).  While we seem to be able to learn from recent black swan events, anticipating the next different black swan doesn’t get any easier.

Ten principles for a black swan proof world

An article in the Financial Times posts these principles from Taleb that addresses economic issues surrounding dealing with black swan events (text selected for relevance to the climate issue):

1. What is fragile should break early while it is still small. Nothing should ever become too big to fail.

2. No socialisation of losses and privatisation of gains. Whatever may need to be bailed out should be nationalised; whatever does not need a bail-out should be free, small and risk-bearing.

3. People who were driving a school bus blindfolded (and crashed it) should never be given a new bus. It is irresponsible and foolish to put our trust in the ability of such experts to get us out of this mess. Instead, find the smart people whose hands are clean.

7. Only Ponzi schemes should depend on confidence. Governments should never need to “restore confidence”.  Simply, we need to be in a position to shrug off rumours, be robust in the face of them.

8. Do not give an addict more drugs if he has withdrawal pains. Using leverage to cure the problems of too much leverage is not homeopathy, it is denial. The debt crisis is not a temporary problem, it is a structural one. We need rehab.

10. Make an omelette with the broken eggs. We need to rebuild the hull with new (stronger) materials; we will have to remake the system before it does so itself. Let us move voluntarily into Capitalism 2.0 by . . . teaching people to navigate a world with fewer certainties.

Anticipating the climate black swan

The title of this post is taken from an article of the same name by Michael Ferrari, who raises the issue of extreme weather event black swans in the context of climate change.

Weather has these Black Swans too. It follows that blindly assuming continued warming climate trends will lead to the observation of slightly warmer temperatures year over year may work for a while, until a cold snap like the one that hit the eastern US at the beginning of last winter produced one of the coldest starts to winter in many years. 

The recent winter weather arguably qualifies as a climate black swan, since people were predicting increasingly warm winters and the disappearance of snow.  Once the anomalous and surprising event occurred, in hindsight the weather was easily rationalized in a variety of ways.  The recent Queensland floods are another example, after the expectation of continued worsening drought conditions with climate change.

In an article by Marc Gunther on Climate Policy and Black Swan,  which is rather unremarkable, there is one interesting quote:

What does this have to do with Taleb? I’m wary of trying to summarize his worldview but Taleb essentially argues that we know a lot less than we think we know. “My major hobby,” he writes on his website, “is teasing people who take themselves & the quality of their knowledge too seriously & those who don’t have the courage to sometimes say: I don’t know….”  In essence, Taleb says we are not very good at understanding the past or present — his first book was called Fooled by Randomness — and that we are downright horrible at predicting the future. 

Joe Romm asks the question “Is global warming a black swan?”  He says:

Global warming obviously is not a black swan.  It is an event “outside the realm of regular expectations” but one can’t say “nothing in the past can convincingly point to its possibility”   In my 2006 post, I argued that rapid polar warming and the potential for a melting of the tundra and massive release of methane was a black swan.  I suppose, for 99% of policymakers and the media it is a black swan, but in fact even the worst-case scenario for global warming isn’t technically a black swan. We have been warned as much as one could reasonably expect us to be warned, but we choose to ignore the warnings.  In fairness, though, there is a massive fossil-fuel-funded disinformation campaign out there trying to convince us that all swans are in fact white.  Would it were so.

Schneider et al. have a relevant paper entitled “Imaginable surprise in global change science“, which was published in 1998 before black swans became fashionable.  From the abstract:

Decisionmakers at all scales (individuals, firms, and local, national, and international governmental organizations) are concerned about reducing their vulnerability to (or the likelihood of) unexpected events, ‘surprises.’ After briefly and selectively reviewing the literature on uncertainty and surprise, we adopt a definition of ‘surprise’ that does not include the strict requirement that it apply to a wholly unexpected outcome, but rather recognizes that many events are often anticipated by some, even if not most observers. Thus, we define ‘imaginable surprise’ as events or processes that depart from the expectations of some definable community. Therefore, what gets labelled as ‘surprise’ depends on the extent to which what happens departs from community expectations and on the salience of the problem. We offer a typology of surprise that distinguishes imaginable surprises from risk and uncertainty, and develops several kinds of impediments to overcoming ignorances.

Overall, the idea of climate black swans hasn’t been explored very much, but I think it is something that deserves further consideration.  I agree that global warming itself shouldn’t be regarded as a black swan, although if some of the more alarming sensitivity predictions were to be true with accompanying extreme weather events, then it might arguably be a black swan.

The black swan concept seems useful in the context of Ferrari’s  thesis, namely related to extreme weather events that are the opposite of what you would expect in a warming world.

Back to the Climatic Change article

So after pondering the black swan-climate nexus some more, I have decided to add the following text to paragraph cited at the beginning of the post:

A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries.  This would entail  understanding and analysis of the regional climate dynamics of extreme events on timescales of decades to centuries.  Such an understanding is needed as the background against which dangerous climate change is assessed.  The unexpected cold, snowy northern hemisphere winter (2010/2011) and the flooding (2011) in drought stricken Queensland highlights how our expectations of extreme events in a warmer world can be soundly trumped by natural variability of weather processes.  Such variability has important implications for the assessment of dangerous climate change and for reducing vulnerability to weather disasters.


465 responses to “Anticipating the Climate Black Swan

  1. Judith, My reading of Taleb’s book is if you can predict it it isn’t a black swan. So trying to prdict potential black swans would be a contradiction in terms.

    He also said that in the financial market that extreme events were not normally distributed and that looking at their probabiilities as if they were lead to poor predictions. Ie. extreme events happen more often that we think. The market crashing would not be a black swan, we know that will happen at some point. The specific cause of a market crash is the Black swan. And it is unknown. If it was known it would be avoided.

    In terms of weather events that would be the same as saying that while rainfall amounts may look normal at firat glance the incidence of droughts and floods would happen more often than a normal model would imply. Hundred year events would hapen more often than every 100 years on average.

    In terms of “climate” having the average temperature take a radical change either up or down would not be a Black Swan. We know that it is going to happen. The Black Swan would be the proximate cause of such a change. I don’t know what a Black Swan event would be, a new ice age, an astroid strike, a shut down of an ocean current…??? None of those has anything to do with the currant CAGW debate nor ar they anything that we can prepare for.

    • Re black swans, it is more about imagining them than predicting them, although one common feature of black swans is that in hindsight, it is usually rationalized that we had enough knowledge to predict this. And my point here is that the climate black swans probably don’t have that much to do with AGW, unless the warming turns out to be extreme. I think last winter’s weather would qualify as a black swan: totally unanticipated (and counter to expectations), with widespread adverse impacts (and in hindsight, we had the information if not the knowledge to predict something like this.)

      • Re black swans, it is more about imagining them than predicting them, although one common feature of black swans is that in hindsight, it is usually rationalized that we had enough knowledge to predict this.

        Indeed. 9/11 was a good example of this. Prior to the event, use of an airliner as a weapon was a plot line in a Tom Clancy novel. After the event, the unconnected dots came out.

      • Pooh, Dixie

        Bullseye! After 9/11/2001, people claimed that the use of airliners was unimaginable. Clancy’s “Debt of Honor” was published in 1994.

        What is now worrisome is the plot line “The Sum of All Fears”.

        Clancy stopped writing novels with such plot lines. I hope he was hired as a consultant by the professional side of the CIA. :-)

      • I am in agreement with JKnapp. “Black Swan’s” are less about being unlucky with a very unlikely roll of the dice. They are really a symptom of a group of people so in love with their mathematical models, they fail to see where their models and nature differ dramatically and dangerously so.

        What was so totally unanticipated about last winter’s weather? There was nothing unusual about it in historial context. People who chose to believe the GIGO computer models might have been surprised. Those who have bought into real Global Warming might have been surprised, but logically why would +half a degree made last year’s “weather” so surprising? People who put their trust into experts (who so believe in AGW) might have been blindsided, but they pay taxes for a service so as not to be blindsided.

        People who said, “Our childern will grow up not knowing about snow.” were the ones going out on a limb.

        “Black Swan” is an excuse for shoddy science. “Nature didn’t fit my model.” My hope is the term Black Swan will soon be regarded with as much scientific respect as “the dog ate my homework.”

      • “Black Swan” is an excuse for shoddy science.

        Not at all. Shoddy science is implicit in assuming that all that we need to know is known (what makes CAGW theory look so much like religion to some). Black swans are quite different. Random, extremely influential events can have apparently disproportionately large effects. Importantly though, the effects frequently, especially since the dawn of MSM, become magnified by reporter ignorance, pursuit of a “story,” even where none exists, and popular fascination with horrible consequences. Rather than information – which “people have a right to know” – we get drama, like the pointless focus on the Fukushima power plant when the real story – and a genuine Black Swan – was the shear power and unusual nature of the earthquake that generated the tsunami.

        A conscientious approach to prediction has to be constrained and qualified. It must explicitly lay out the assumptions that must be correct for the prediction to be true. “If our models are good” does not meet this requirement. Even “if the assumptions upon which our models are based are correct” misses the necessary caution.

      • Duster, I have considered your post. I also read, “Black Swan or Black-Scholes: The Crisis of the Actuarial Profession.” (Smedinghoff, Contingencies May/June ’08)
        http://www.contingencies.org/mayjun08/swan.pdf

        This is an approving critique of Talib’s “Black Swan”. Smedinghoff says that this book should be “on the list of forbidden works among actuaries. Ideally, The Black Swan would have an immediate and permanent
        place on the Society of Actuaries exam syllabus. But the resulting cognitive dissonance could be too great for most actuaries.
        This is because Actuaries live by the Gaussian distribution, but the real world, where it matters, doesn’t always play by those rules.

        I’ll amend my statement:
        “Black Swan” will become an excuse for shoddy science unless we put our foot down.

        By your own argument, you should “explicitly lay out the assumptions that must be correct for the prediction to be true.” That is a necessary, but not a sufficient condition for good science. The assumptions you use must also cover the real world domain of influences and forces. It might require messy Weibull with messier cross-correlations for it to yield useful predictions where it really matters. Making the model tractable does not necessarily make it closer to reality.

        “Nature did not fit my [far too simple] model. Therefore the missed prediction was because of a Black Swan event.” Twaddle. That will be just an excuse for not doing proper science to build a model that can predict future events. If the science is too complex for that, then you must realize you prediction is not worth much in the first place.

        “Black Swan” must be understood as “I screwed up big time.”

      • Stephen
        My hope is the term Black Swan will soon be regarded with as much scientific respect as “the dog ate my homework.

        Please don’t malign the dog. It was actually our pet goat who ate my son’s homework….and the architect’s plans, the builder’s tobacco, my wife’s apple turnovers and the neighbours’ roses.

      • These could be regarded as ‘black goat events’, perhaps the opposite of ‘black swan events’ in that the former were entirely predictable and potentially preventable.

      • There’s also the issue of regional clumping, to coin a phrase. Within a regime of “average”, there can be intense concentrations of certain phenomena that are not only remarkable, but have knock-on effects that a smoothly averaged prediction can’t capture. E.g., getting half a year’s average precipitation in a week, in an area 1/10 of the prediction cell size. .

      • But weren’t last year’s broad trends for weather predicted, IIRC over at WUWT by Joe Bastardi, using historical observations only? It seems that the confluence of La Nina and several other cycles did actually perform to somebody’s expectations.

        I’m not 100% sure of the above and hope somebody fact checks me quickly.

      • Tom –
        Bastardi did make predictions and, in fact, there’s a Youtube video out there someplace. But, IIRC, those predictions were for more than just last winter.

        And after looking for it, there are several of them. This one was what I remembered –

      • The weather of this winter was not a surprise to everyone, only those with a bad case of climate science hubris. The recent financial crisis was due to models which assumed that real estate prices cannot go down. That may have made it look like a black swan to idiots, the historically illiterate, and Wall Streeters blind with runaway hubris. But normal people know that asset prices can fluctuate.

        The important point you should take away from Taleb is that experts are not nearly as smart as they think they are. See, e.g. Teltlock.

        And for a great explanation of how group wisdom should be gathered (contrary to the IPCC which did everything wrong) see the Wisdom of Crowds.

        The experts aren’t any better at predicting the future than a chimp throwing darts.

      • Oops, sorry. Meant to point out that one of the points made repeatedly in the Wisdom of Crowds is that the crowd performs much better than the experts far more often than not. Anyone curious how a bunch of bloggers (and their armies of davids armed with all manner of expertise, knowledge, experience and wisdom) can consistently run circles around the hockey team’s self-appointed “experts” of climate science needs to read the book. History is chock full of examples.

      • A black swan is not just an un-anticipated event. It is an event that could not reasonably been anticipated.

        For hundreds of years it was widely understood that swans were white. There was no lack of research or lack of clear thought — overwhelming evidence forbade the existance of a black swan. When one was sighted (in Western Australia) it turned accepted science on its head.

        Severe whether of last year was not anticipated, but was not considered impossible. An example would be if it snowed in Death Valley in June, while the air temp was 110 degrees F. This would indicate that we had fundamentally misunderstood the science.

        While you cannot anticipate a black swan, you can acknowledge that something completely unexpected might happen (though you don’t know what) and try to plan with sufficient flexibility to be able to weather the unexpected.

      • Maximizing available resources and flexibility is the only ultimately useful strategy. The lack of either or both is what guarantees maximum disruption.

      • David L. Hagen

        Judy

        How do you categorize the inevitable decline of light oil?
        The 800 lb gorilla?
        (It has been predicted and is expected by quite a number, so neither black swan nor dragon king.)
        The only issues are the timing, and rate of decline until mitigation can be brought on line.
        That requires far greater attention than AGW.

        See Robert L. Hirsch, The impending World Energy Mess.

      • Anent which, Romm’s comment and claim to be right about tundra melting etc. is pure bumpf. The climate black swan we’re in need of imagining and looking out for is continued or accelerated cooling. That’s megadeath territory, unlike the nonsense spouted by Gores and Romms.

  2. Sorry for the typos.

  3. The recent winter weather arguably qualifies as a climate black swan, since people were predicting increasingly warm winters and the disappearance of snow.  Once the anomalous and surprising event occurred, in hindsight the weather was easily rationalized in a variety of ways.  

    AR4

    The ensemble mean of the MMD models projects a general decrease in snow depth (Chapter 10) as a result of delayed autumn snowfall and earlier spring snowmelt. In some regions where winter precipitation is projected to increase, the increased snowfall can more than make up for the shorter snow season and yield increased snow accumulation. Snow depth increases are projected by some GCMs over some land around the Arctic Ocean (Figure S10.1) and by some RCMs in the northernmost part of the Northwest Territories (Figure 11.13). In principle a similar situation could arise at lower latitudes at high elevations in the Rocky Mountains, although most models project a widespread decrease of snow depth there (Kim et al., 2002; Snyder et al., 2003; Leung et al., 2004; see also Box 11.3).

    Figure 11.13. Percent snow depth changes in March (only calculated where climatological snow amounts exceed 5 mm of water equivalent), as projected by the Canadian Regional Climate Model (CRCM; Plummer et al., 2006), driven by the Canadian General Circulation Model (CGCM), for 2041 to 2070 under SRES A2 compared to 1961 to 1990.

    #####

    The recent Queensland floods are another example, after the expectation of continued worsening drought conditions with climate change.

    AR4

    A range of GCM and regional modelling studies in recent years have identified a tendency for daily rainfall extremes to increase under enhanced greenhouse conditions in the Australian region (e.g., Hennessy et al., 1997; Whetton et al., 2002; McInnes et al., 2003; Watterson and Dix, 2003; Hennessy et al., 2004b; Suppiah et al., 2004; Kharin and Zwiers, 2005). Commonly, return periods of extreme rainfall events halve in late 21st-century simulations. This tendency can apply even when average rainfall is simulated to decrease, but not necessarily when this decrease is marked (see Timbal, 2004). Recently, Abbs (2004) dynamically downscaled to a resolution of 7 km current and enhanced greenhouse cases of extreme daily rainfall occurrence in northern New South Wales and southern Queensland as simulated by the CSIRO GCM. The downscaled extreme events for a range of return periods compared well with observations and the enhanced greenhouse simulations for 2040 showed increases of around 30% in magnitude, with the 1-in-40 year event becoming the 1-in-15 year event.

    • AR4 sure sucks when it comes to ability to predict climate.

      • While this may eventually be the case, I’m scratching my head wondering how you could make such a statement only a few years after it was written.

      • This is easy. Get the next 3 in a row wrong, and simple probability says you suck at predicting.

      • IPCC is 21 years old at least. AGW is supposedly 40 years old. So a prediction based on AGW should certainly be correct by now if they knew what they were talking about.

        Instead it appears climate is cyclic and matches up better with the PDO/AMO etc than with CO2.

      • Um… did you guys get that AR4 correctly predicted the weather conditions that the original post described as “anomalous and surprising?”

        In other words, that they weren’t at all “anomalous and surprising,” but expected, having been predicted four years previous?

      • I forget sometimes that the IPCC has been so distorted over time that not even the actual words from the document can stop the momentum of proving it wrong, post haste! You need to start with the assumption that all IPCC scientists are either corrupted or idiots, then not care about getting it at all right. This seems to cover the bases.

    • grypo,
      So you are happy with the ‘if it rains, droughts, snows or not, blows or becalms, it is all AGW?
      If one reads the quotes, in context, of what you are offering, all it says is that anything that happens, good bad or indifferent is attributed to CO2.
      Sorry, but that sort of garbage works in certain faith traditions, but is useless garbage in terms of science based policy.

      • This is just a broader distortion of the predictions than the initial post was.

      • grypo,
        I apologize for mis-characterizing what you were trying to say and for being short with you in my reponse.
        I should have read you more carefully.

      • grypo,
        No, you just don’t like the implications of my question so you are dodging it.
        The IPCC, as the marketing arm of the AGW consensus, is supplying all of the distortion required.

    • Last year was a year of Low Arctic Sea Ice Extent. That is always followed by a winter with extreme Snow and Cold. NOAA does not Forecast these events because it does not fit with their Theory and Models. It does fit with the history of what happens. Some of us did expect the Snow and Cold because this always happens. When the Arctic is thawed, it will snow and be cold during the following winter and that does cool the earth.

    • John Kannarr

      AR4 sounds like a typical horoscope, able to be construed to match almost any subsequent occurrences of weather.

      • Are you suggesting that “a tendency for daily rainfall extremes to increase under enhanced greenhouse conditions in the Australian region” can be construed to mean “the expectation of continued worsening drought conditions with climate change?”

        If so, please explain how that passage can be so construed. It seems unambiguous to me, but I haven’t looked at horoscopes since I was a boy.

  4. “The unexpected cold, snowy northern hemisphere winter (2010/2011) and the flooding (2011) in drought stricken Queensland highlights how our expectations of extreme events in a warmer world can be soundly trumped by natural variability of weather processes.”

    It seems to me that the 2010/2011 winter and the Queensland flooding were only black swans to those whose “imagination” was constrained by their fervid belief in AGW/CAGW. Both types of events, severe early winters and flooding in Queensland, have happened repeatedly in the not so distant past. The use of the word “our” with respect to expectations, may be a bit over inclusive. Not everybody was so shocked by these events. This lack of vision seems to be more a sort of self imposed blinders than an example of the limits of human knowledge or imagination.

    Is a black swan still a black swan in this sense, when there have been birds of a similar hue in the flock in the recent past, but the ornithologists have concocted a theory that there can be no more black swans in the future? Or are they just evidence that the ornithologists are just, perhaps, wrong?

    • kalsel3294

      I have to agree entirely with GaryM regarding the Queensland floods.
      It was the coinciding of specific conditions in the Indian and Pacific Oceans that created the conditions this time around, just as the same coincidence had done so in the past.
      The difference now is that some researchers are close to reliably predicting SST’s 2 years ahead.
      However the range of various predictions from various agencies is so wide that as a whole they are virtually useless, but there are one or two who are establishing a good track record.

    • Agreed, the 2011 floods were consistent with the experience and patterns of the last 150 years, they would surprise only those who seek to project into the future on the basis only of the very recent past. Brisbane City Council (and others) had been anticipating a major flood since October 2010, the severity of the flood was related to bad dam management as well as high levels of precipitation. Much of the flood damage in several areas occurred in part because political/economic interests in development overrode the concern of relevant professionals about flood-prone areas.

  5. Interesting, and thanks for posting this. One problem……..
    “This would entail understanding and analysis of the regional climate dynamics of extreme events on timescales of decades to centuries”.
    My question is do we understand climate to the point where we know what is natural variability of weather processes? At what level of weather activity do we attach the word extreme? Who makes this choice? I doubt we even understand the basic concept of global norms. We study climate inside a microscopic time frame and assume that we can now understand 4 Billion years of global climate. Then again, I’m not a scientist. ;) I just like to read.

  6. Couple of things
    [1] “A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries…”.

    It seems to me that we knwo enough about natural climate variation that these extreme weather events are expected. I remember ice climbing (ice climbing!!) in North Wales during the winter of 78 in -18 deg. C. In that context recent winters are not altogether “unexpected”. The explanation of them (warm oceans cold continents weather patterns) is just somewhat more recent. So are they really black swans?

    and “The unexpected cold, snowy northern hemisphere winter (2010/2011) and the flooding (2011) in drought stricken Queensland highlights how our expectations of extreme events in a warmer world can be soundly trumped by natural variability of weather processes”.

    I am confused by this sentence. On the one hand it appears to say that greater extremes are expected but then that those expectations are trumped by natural variability – it seems like a contradictory statement that needs some clarification.

    • Black swans need not be totally unexpected. They must only be rare enough that most people ignore their possibility and big enough events to be significant in spite of their rarity. Hurricane Katrina was certainly one, because it hit New Orleans and caused all the damage. The very recent tornadoes might also be considered a black swan due to the large death toll. The 2010 heat wave and related forest fires in Russia might also be considered a black swan.

      For climate a major shift of decadal influence – in either direction – could be considered a black swan, and the realization of a tipping point would certainly be one.

      • Do not confuse Black Swans with Complacency. A Cat 5 hitting New Orleans was always recognized as a danger, but people chose to live with it. Shelter in the Super Dome; what could go wrong?

        Fukushima might qualify as a Black Swan where all the desiel backup generators are ruined in the Tsunami following the quake. Might. Do you want to bet in 40 years no one suggested another independent set of backup generators higher up the hill might not be a bad idea?

      • More on alleged Complacency. How about barges in the wrong place at the wrong time? Like ING-4727 in or near the Industrial Canal. http://en.wikipedia.org/wiki/ING_4727

      • We may all have our own ideas on what makes a black swan.

        Reading the book of Nassim Nicolas Taleb, he is not very restrictive on what is required. His major examples include a stock market crash. Those are certainly not unforeseen at the level that crashes occur, but they are unforeseen at the level of timing and specific details. In this respect they are just like hurricane Katrina.

        The impression that I was left on the content of the book “The Black Swan” is that most of the examples belong to the same class, but I haven’t checked (and do not plan to check), whether this is really the case.

      • To be honnest, fukushima is not a blackswan, but a dragonking.

        the blackswan was the M9 earthquake, but even that is not so much a BS.
        the cause of the accident is the fact that the geopgysicist community (source SciAm in french version) in the 70s was feelin that this could not happens where it happend, because the rift/break (correct me, I’m alien) was supposed to glide nicely with small earthquake, and quakeless glide(assumed because ther was difference between earthequake constraints relaxed, and total constraint created by usual move)…
        after that, people discover that ther have been big tsunamis, big earthquake, and that the theory may be false…
        but the power plant was built on a pretend safe place, and not beeing sure of the new theory, not much was done to anticipate a huge tsunami.

        after that the M9 quake, cause a more that 10m tsunami, thus break all diesel, kills thousands of local, destroy road and infrastructure, smash the mindset of the workers and the dynamic of the governement…

        a blackswan in the blackswan vs dragonking theory is not an unexpected only… it is a chain of improbable failure that propagate, but not so improbable because the farther it propage the most probable it is to propagate further…
        it that theory it ils linked to a group of failing point that are quite different but connected. everyone have a good propability to break the propagation, but the longer is the chain, the harder it is to stop…

        the typical blackswan is an earthquake, an avalanche, a buble explosion in a varied marked… it is typically fractal, self similar…

        a dragonking is not a surprise for God.
        the failure of the agents is caused not by a chain of failure that propagate, but by a failure caused by the same cause… a unique way to think, a common failure point.

        you can find dragonking in centralized organizations, in unified mindset (like modern financial markets with brainwashed actors and robots that trains them)

      • Katrina was expected, not by name or date but certainly by event. I recall my reaction when I heard of it, namely that it had finally happened. One only had to stand downtown in New Orleans and look up at the river boats passing by to expect it “someday.” No one should be surprised when a city below sea level goes under. Shocked sure, but not surprised.

        But then surprise is not a scientific concept. Perhaps we are merely shocked by black swans. Since we are debating a metaphor anything is possible.

      • John Carpenter

        I think a black swan has to be an unforeseen event rather than a rare, but possible event. Anyone living in the Carribean, Florida and Gulf coast and even upwards to Long Island and Cape Cod are aware category 5 hurricanes can come… because they have in the past. There is precedence. IMO a black swan has no precedence because “all swans are white”. Therefore the appearance of one is totally enexpected. Tornadoes, hurricanes, droughts etc… all occurring in places where they have occurred before are not really black swans… we know they can happen and they will and do happen… so this should be no surprise when it does happen.

        In the context of climate change, my idea of a black swan would be the climate getting colder again… and staying colder. At this point in time, I would guess most climate scientists would only consider this to be a crank theory. Most everyone in climate science agrees there will be more warming to one degree to another, so even greater warming than expected would not really be a black swan either as some climatologists have predicted as much. I am not aware of any serious research indicating a coming long term (more than 20-30 years) cold period on the horizon.

        If natural variation provided such a thing to happen… from out current knowledge perspective, that would be the black swan.

      • John Kannarr

        What about a new type of weather event never before recognized or experienced? That would be a true black swan. Perhaps the modelers should be searching the local effects (i.e., weather) within their models for such unexpected situations, assuming anyone actually believes the models have any validity beyond scare propaganda.

      • the way i understand it, something not recognized or experienced would be a dragon king. black swan is something on the tail of the distribution

      • Dear Judith

        This is the abstract or part thereof of Sornette 2005.

        ‘We emphasize the importance of understanding dragon-kings as being often associated with a neighborhood of what can be called equivalently a phase transition, a bifurcation, a catastrophe (in the sense of René Thom), or a tipping point.’

        As everything in Earth systems is chaotic – Dragon Kings are the extreme but meaningful outliers that occur at phase transitions. The El Nino in 1998 for instance.

        9/11 was a black swan – precisely because it was not predicted. It was not part of a sampling of events that could have a distribution – although the impact was large. Whereas the death of Osama was predicted – a well deserved justice. It was a one off event hopefully.

        I think you may be confusing the narrative.

        Cheers

        Cheers

      • Chief –
        Actually, 9/11 was not unpredicted – it was simply ignored as being possible. There was at least one and, I believe, two books written prior to the event where that mode of attack was the central theme.

        I think what many people are missing here, although it’s been said by a few, is that Black Swans are not generally unforeseen by everyone, but are, rather, surprises to individuals or to some specific group(s). 9/11, for example was certainly not a surprise to the perpetrators although the success of the effort evidently was. The recent financial crisis was not a surprise to everyone although it certainly was to many. An eruption of the Yellowstone caldera would be a Black Swan to the many who know nothing of the possibility, but not to those who understand that it’s possible. IOW, it may be an individual Black Swan – or it may be a group Black Swan, but only something like the overnight appearance of interstellar war on our planetary doorstep is likely to be a universal black Swan. And even that has long been considered in science fiction – and more recently by Stephen Hawking.

      • Jim,

        I did consider my words there carefully. The definition of a black swan includes a caveat that it applies to an observer.

        Cheers

      • CH, Osama’s death will definitely be a one-off event.

      • Hi, Judy. The way I’m reading that, I think you have the two reversed. Sornette calls dragon kings “meaningful” outliers, that “coexist with power laws.” So something like the recent earthquake in Japan – though large and unusual – would be a dragon king. (I should mention I’ve not heard the term ‘dragon king’ before now.)

        I think a black swan event is more like the failure of the Fukushima nuclear plant, where a number of unlikely events had to conflate (earthquake + tsumani + failure of the surge wall + backup generators in the basement) to cause – in this case – a disaster.

        I may have to re-read Taleb, but the main example I remember comes from the first chapter, regarding WWI. There was a series of unlikely events that led to the assassination of Archduke Ferdinand, which further led to the war. This wasn’t on the tail end of a distribution, but was a one-off: both explainable in retrospect and inherently unpredictable.

        Regardless of all that, I think the addition you’ve made in this section of your paper is a good one.

      • Hi Ted, read this little article entitled “How dragon kings could trump black swans”
        http://www.technologyreview.com/blog/arxiv/23935/

      • That article seems to be again a case of overextending some rules-of-thumb as most power laws are really rather rules-of-thumb than anything deeper. They are not accurate, they are not valid over the whole range from zero to infinity, but there is always a cutoff somewhere and a failure at the other extreme as well.

        Very few power laws are really based on well defined mechanisms. The power law is just a flexible enough concept to allow for a reasonably good fit to a part of many distributions which have much fatter tails than the Gaussian distribution.

        They are a good rule-of-thumb as the logarithmic dependence is a good rule for the relationship between the CO2 -concentration and forcing over a rather wide range of concentrations.

      • (from ref. 23_935)
        Sornette looks to be on to something interesting with his notion of dragon kings: outliers that exist beyond the usual realm of power laws. That could be a hugely influential. But his contention that these outliers are in some way more easily predictable than other events smacks more of wishful thinking than of good science.

        From Sornette 2009:“An outlier is an
        observation that lies an abnormal distance from other
        values in a random sample from a population.” It is therefore an anomaly, an event to be removed in order to obtain reliable statistical estimations. The term “outlier” emphasizes the spurious nature of these anomalous events, suggesting to discard them as errors, or as misleading monsters.
        (Sec. 3)
        It was here I resolved to consign “Dragon King” into the dust bin of hair-brained ideas. You remove the most important data from the distribution you are trying to represent? Here’s a better idea: don’t throw out your important outliers; you throw out your candidate distribution as a poor fit and try something else. The Sornette paper only gave superficial reference to Black Swan Theory, hardly enough to be justified in using the phrase in the title.

        In Sornette’s first example, he says Paris is a Dragon-King in the power-law distribution of city sizes. May I point out that Paris is a known, yet is still classified as a Dragon King. Suppose we hide Paris and try to predict it? It is an artificial test (use sizes 2 thru 220 to predict #1), and it will misestimate the size of Paris because it is a bad model to predict the largest anything.

        If his Fig 15 isn’t a poorly chosen distribution, on an ill fitting model, what is? And the remaining few [DK] events have been found to be statistically different: the hypothesis that they belong to the same distribution as 99% of the population of the other drawdowns is rejected at the 99.9% confidence level [32,33]. Yes. Where is it written we are sampling from only ONE Power law distribution and not a mix of two or more?

        The concept that there are non-linear results to consequences is well understood in engineering. Material Yield Points and Fatigue are real. When systems approach a fail point, behavior becomes very non-linear. It was only 4deg C colder than the previous coldest day for a Shuttle launch when we lost Challenger. Results beyond the domain of your control and training data points will be full of surprises. “Here there be dragons.”

        But we don’t need Sornette’s new concept of “Dragon King”. A Black Swan is an event whose probability of occurrence was underestimated for a variety of reasons. Dragon Kings are extreme events [which] occur much more often than would be predicted or expected from the observations of small, medium and even large events. I can only conclude that Dragon King is a subset of Black Swan and does not warrant a separate name nor distinction.

      • as I explain earlier,
        for me fukushima accident is not a chain of problems, but the normal consequence of the tsunami. it is a dragonking…
        the huge consequence of a fact, because we ignore that many supposed independent facts, get correlated by an unknown factor…

        note that my previous comment, ignore what is explained in another comment. that a chaotic system my exhibit hidden strong correlation when changing phase, leading to huge unexpected results…

        imagine the surprise of a bird when water get solid in a cloud… hard to anticipate such a change if you never experience it…

      • I believe the concept of the dragon king is just recognition that statistical outliers should be expected among any natural set of data. So I believe John was correct that a new kind of weather event would be a good example of a black swan–a kind of swan not seen before. A dragon king would be something like a cat 5 hurricane.

      • Harold Pierce Jr

        You should check:

        “Climate Change and Long-term Fluctuations of Commercial Catches: The Possibilty of Forecasting” by K.B. Klyashtorin

        FAO Fisheries Technical Report. No. 410. Rome, FAO. 2001

        Available at:

        ftp://ftp.fao.org/docrep/fao/005/y2787e/y2787e00.pdf.

        By analyzing climate data and fish catch data, Klashtorin found the earth has a general climate cycle of 55-60 years.

        Note date of publication. Was this report passed onto the IPCC?

        “Cyclic Climate Changes and Fish Productivity” by K.B. Klashtorin and A.A. Lyubshin, which you can download for free thru this link:

        http://alexeylyubushin.narod.ru/Climate_Changes__and_Fish_Productivity.pdf?

        NB: This mongraph is 224 pages. This book is not about climate science. The Russian edition was published in 2005. The English translation was published in 2007 and was edited by Gary D Sharp.

        By analyzing a number of time series of phenomena influenced by climate, they found that the earth has global climate cycles of 50-70 years with an average of about 60 years and which have cool and warm phases of 30 years each. They summerize most of the studies thru early 2005 that show how this cycle influences fish catches in the major fisheries.

        The last warm phase began in ca 1970-75 (aka the Great Shift) and ended in ca 2000. The global warming from ca 1975 is due in part to this warm phase. A cool phase started in 2000, and their stochastic model projects that it should last about 30 years. See Fig 2.23, p 54.

        See also Fig. 2.22 (p. 52) and Table 2 (p. 53). They show that increasing world fuel consumption (i.e., increasing CO2 emission) does not correlate with cool and warm phases of the 60 year global climate cycle. That is to say, they show that increasing CO2 concentration in the air does cause global warming.

        Was the IPCC aware of this seminal monograph and the climate projections made by the authors that are in conflict with their projections of warming?

      • I think that in your penultimate sentence you meant to say that “increasing CO2 concentration in the air does not cause global warming.”

      • We have had Tornadoes and heat waves before. The Russian’s had reserves of grain that helped them through the crisis. Has a major shift of decadal influence/tipping point ever happened before? Maybe the little ice age which people survived, probably without even being aware what it was, but has there been one for heat? correct me if I am wrong on that. If not, the consequences are unpredictable. It’s hard to prepare for if it has never happened before.

        We also survived the event of a black swan okay too! :)

        What exactly do you mean by tipping point?

      • A tipping point as Sornette used it refers to a chaotic bifurcation in complex dynamical systems. Smaller bifuractions in Earth systems occurred around 1910, the mid 1940’s, the late 1970’s and 1998/2001.

        In the Holocene larger bifurcations around 12,000 and 8,000 years ago – significant and very rapid warming and cooling. There is evidence in sediments for an ENSO bifurcation 5000 years ago that resulted in the drying of the Sahel.

        The preparation for any and all of this is to build strength and resilience in human cultures and economies.

      • Pekka –
        Wrt Katrina – there were several things that could be called Black Swans, but the hurricane wasn’t one of them. That was expected – and, in fact, was less than a Cat 5 when it hit the coast. The real Black Swans were 1) the failure of the sea wall (which should have been considered but wasn’t) and 2) the governmental failure at both the city and state level. FEMA performance was certainly not optimum, but that was, in part, because the LA Governor failed to request the aid that was waiting for that request to be made.

      • Arguing on a metaphor is fruitless. Even Taleb cannot decide, how others must understand the concept.

        The real message of his book is in my understanding that the total importance of all possible rare events is large. It’s so large that describing the world without their contribution would lead to seriously erroneous conclusions. Their frequency is commonly underestimated. There is no reason in this connection to separate rare coincidences of more common contributing factors from events due to single rare origin.

        Benois Mandelbrot is referenced in many places by Taleb. All these references are related to his interest in probability density distributions of fat (power law) tails. Such events are known well enough to define PFD’s, but they form also an important part of the discussion of Taleb.

      • In short, models which can not take account of extreme events can at best be only a general guide as to what might happen, and policymakers must be aware that they may well be seriously inadequate. This is commonly seen in economics. CGE (computable general equilibrium) modelling is a good tool for understanding relationships and assessing the relative costs and benefits of policy A versus policy B (and they would generally be restricted to ten years rather than the IPCC’s 100). But economic forecasts are notoriously poor, as they of necessity are based on continuation of existing trends. I would never base economic policy advice purely on modelling outcomes, I would think that would apply even more forcefully in the case of climate models involving much more complex and less understood systems.

        My advice re climate, as with economics, is to foster the community’s capacity to respond positively to changing circumstances, whatever they may be. This is a far superior strategy to focusing policy on one predicted or projected outcome in an uncertain world.

      • Yes. A major theme in Taleb’s book was that of serendipity. The idea also speaks to discovery and having an open mind.

        In regards to serendipity, the book says you must be ready to take advantages of windfalls. I haven’t seen much news regarding how much better a warmer planet would be (longer growing seasons for one). For the unpredictable disasters we adapt when it comes along. Cross the bridges when we come to them.

  7. The Black Swan in this is that there is no climate calamity taking place.

    • hunter

      What’s “black” about that “swan”?

      Max

      • manacker,
        The black swan is the existing reality that the experts decide is not there.
        The climatocracy has decided that this outcome, that things are not changing significantly, is one that does not exist.

      • When an expectation is created, and not fulfilled, it leaves the experts scratching their heads and looking silly. Remember the Y2K bug? What was the black swan, the catastrophe that didn’t happen, or the nothingness that did?

      • ChE,
        Y2K brings up a great point.
        Many thought it was over blown, and it seemed to have fizzled out pretty well by mid-year of 1999.
        I guess in context of where I see this blackswan going irt things presented as pop-hysteria, the black swan is the swan that didn’t trumpet.

      • Y2K was mostly fanned by government. I was consulting to government agencies at the time, and made a bundle by just making lists of equipment, calling the vendors, and asking them if they were “Y2K compliant”. Of course there is/was no such thing, but, the clients didn’t know that.

        Also did a ton of generator projects, most of which weren’t finished until after Jan 1, 2000. That was a good year for the generator business, too.

      • Y2K was over-hyped, but real.
        however the failure was very badly popularized.
        the problem was not like a hollywood film with the death of all at 00:00
        the risk was not of plane crash, nut morning opening, end of day, end of month, end of year, bugs in banking, few but real annoyance of infrastructure like doors, lights, … not in one bank, on one random period, when they change their softwares, but at about the same period…

        the same kind of bug happend because of windows version update, hopefully it is done over one or 2 years…

        to be honnest there was Y99 bug the year after that we discover by the way… and Y01 also… it was the occasion to clean many bugs. but Y00 was much more common and could, if present in many different system cause a big surcharge…
        a kind of dragonking (upset clients, tired workers, diverses problems, hard to find consultant, holtine broken, long delay to solves,impact on road traffic, higher electric consumption, avalanche of problems and impacts, leading to sensibility to small unrelated problems like a power plant break, car accident or a meteorologic problems)

        Y2K correction, were like immunization. it did not save everybody, but it avoids contaminations and epidemia.
        the few bugs that remains, have been solved quietly…

        new versions of windows, evolution to IPV6, have much more impact, but we can choose the dates to break our IT infrastructure one by one.

      • The unfounded hype was over embedded applications, which usually don’t even have a real time clock. This is analogous to the overhyped concern over hackers taking down the electric grid.

        In fact, what was remarkable about Stuxnet was that they were able to use malware to effect an embedded (PLC) controller, though doing it required a permanently connected Windows PC. It doesn’t actually attack the PLC, it zombifies the SCADA PC, which then turns the PLC into a ventriloquist’s dummy.

        Too many people think that Windows PCs run everything, and frankly, too many Windows PCs are involved in infrastructure. But the Y2K panic was completely overdone, and it was obvious to people involved in embedded control.

  8. A potential Black Swan in my view is that increasing CO2 has a very small impact on global temperatures. One of the features of Black Swans as promulgated by Taleb is that they are generally not predicted by experts. Experts and their models turn out to be too simplistic for the chaotic system that they attempt to model, and so fail to predict the Black Swan event. Taleb was talking about economics rather than climate, but I think that there are good parallels here. Should it turn out that the predictions of significant global warning from CO2 turn out to be false, I think it will be easy in hindsight to construct a narrative that showed it was predictable.

    I guess catastrophic global climate change is potentially another – not predicted by the experts, but in hindsight a narrative would be easy to construct.

    • Exactly.
      The black swan had no problem with its exisetnece except in the minds of the finest academics of the day.

    • Mark S,
      Oops, I did not read through to your finishing sentance.
      Please note that we are here precisely because the AGW consensus ahs decided we are facing a “catastrophic global climate change “.
      The black swan would be the unexpected.
      The only thing unexpected in the consensus of today is that weather/climate do what it is in fact doing: nothing particularly unusual. Weather is as dangerous today as it was 100, 150, 200, 500, and more years ago.

      • At the moment, the IPCC “consensus” doesn’t include the scary catastrophic predictions that happen. For example, we are getting 3mm/year sea level rise, but a small(ish) number of scientists are predicting the tens of meters or worse. It doesn’t include the gigatons of Methane that some think might be released from Siberia. We’ve all seen the scary predictions get prominent placement in the media, but these are quite a long way from what AR4 (and previous) has said.

      • Mark S,
        If we were only dealing with the many problems of the IPCC, we would not be in the hole we are in.
        The fear mongers use the IPCC as a starting place to tell us how the victims of last week’s tornadoes deserved it, and how brown recluse spiders are advancing because of AGW and how Manhattan is going to be underwater by last year with a sub-tropical climate.

      • “the fear mongers use the IPCC as a starting place to tell us how the victims of last week’s tornadoes deserved it”

        I think it is exactly the opposite – as the IPCC specifically identifies potential climate problems in order to clarify the situation for policy makers and thus the idea behind the IPCC is to identify and avoid such disasters, then the fear mongers (as you call them) are “mongering fear” (using the IPCC reports) to AVOID climate problems exactly because NOBODY deserves to die – why else do you think they want to mitigate climate change?
        I assume these people don’t hate the victims or hate countries that have high emissions, but that they hate suffering cause by climate change.

        http://mitigatingapathy.blogspot.com/

      • paul,
        No one is mitigating anything in the climate and no one is likely to do anything different for a long, long time.
        Not one policy pushed by the climatocracy has worked in any measurable way except that of financial benefits to those pushing the policies.
        Your closing conclusion begs the question:
        What suffering caused by what climate change?

    • Willis Eschenbach

      Judith, very interesting, you go. Keep up the good fight.

      One issue that deserves greater emphasis is that weather events and observational datasets (e.g. temperature, pressure) do not follow a “normal” (Gaussian) distribution. Instead, they are “fat-tailed”, with an excess of both large and extremely large events.

      In nature, “black swan” events are so predictable that has led to the term the “Noah Effect”. This postulates that in a given set of rainfall records, there will be one record which dwarfs the remainder (Noah’s Flood).

      These datasets are not Gaussian (normal) distributions. Instead they follow some kind of Exponential/Power/Zipf/Pareto/Weibull type of distribution. All of these distributions have an “excess” of large and rare events over the Gaussian distribution. And because of the fractal nature of climate phenomena, this is true at all temporal scales from hourly to millennial.

      So. From my perspective a “black swan” event is an event which is fairly probable under the true distribution of the underlying data, but highly improbable if we assume a Gaussian distribution.

      If I were to study the subject, which I haven’t, I’d start by seeing the overlap at the three sigma level (three standard deviations). I’d consider events that were less than three sigma in the true distribution, and greater than three sigma in the normal distribution.

      Then I’d likely define a measure of “blackness” as being the difference between the true odds and the Gaussian odds of a particular occurrence.

      But surely someone has done all of this, no? In any case, Judith, important issues.

      Next, Mark, I like the idea that:

      MarkS | May 2, 2011 at 4:38 pm

      A potential Black Swan in my view is that increasing CO2 has a very small impact on global temperatures.

      That’s the big one, all right.

      You then say:

      … Should it turn out that the predictions of significant global warning from CO2 turn out to be false, I think it will be easy in hindsight to construct a narrative that showed it was predictable.

      Hindsight? I’ve said since day one that CO2 doubling was a third-order forcing. Total downwelling radiation at the surface? About half a kilowatt per square meter (global 24/7 average solar + longwave).

      Given that half kilowatt, a doubling of CO2 (~ 4 W/m2) is less than a 1% change in total forcing … and in a huge system like the climate, it will be hard to even extract a signal that small, much less attribute it to anything. If a 1% change in forcing were enough to drive the planet to Thermageddon, it would have happened long ago.

      w.

      • Willis, is it possible that Black Swan describes the observers more than the observed? WRT the financial crisis, in retrospect lots of fairly obvious warning signs were pointed out as being available for inspection that got overlooked. Your point about non-Gaussian distributions seems to reinforce the idea that hindsight shows that Black Swan events not only could, but should have been predicted as possible outcomes.

      • Tom,

        Just so. From the Wikipedia entry on Black Swan Theory:

        Taleb states that a Black Swan Event depends on the observer. For example, what may be a Black Swan surprise for a turkey is not a Black Swan surprise to its butcher; hence the objective should be to “avoid being the turkey” by identifying areas of vulnerability in order to “turn the Black Swans white”.

      • Tom –
        The financial crisis was not unforeseen by some of us who were paying attention. The extent of the crisis was not anticipated – nor the continuing effects – nor what I consider to be the inept handling thereof. But the “fact” that a crash was imminent was obvious if one looked only at the housing market and the job situation. 25% increase in housing values per year is not a sustainable situation when income is increasing at (maybe) 3%.

      • Although bizarrely, many people – in fact most people in the industry – did believe that it was sustainable. And they could cite figures to back up their position. Witness the 10% growth per annum in China’s GDP, and witness all those people predicting that China’s GDP will be larger than the US’s within a small number of decades. As if that’s going to happen by 2025 or 2050. But it seems to be prevalent even so. The people closest to the numbers generally do not make the best predictions.

      • MarkS –
        The people closest to the numbers generally do not make the best predictions.

        That’s true in any field you care to name.

      • The same mistake (along with a number of others) were made in the financial models. Risk isn’t normally distributed. And of course, volatility (the substitute academics use to model risk) isn’t really risk at all.

      • A Log-Pearson type III distribution is probably most commonly used.

      • Hi Willis,

        There are always a few who predict the black swans. Read Michael Lewis’s “The Big Short” about the handful of guys who predicted the meltdown of the sub-prime mortgage market, and who profited massively on the decline by effectively taking bets with people who believed the consensus view that the market was going just fine thanks. So if it turns out that AGW theory is overblown, you and a smallish bunch of people who stuck their necks out will be vindicated, and the general public will be wondering how they missed what was so obvious according to the narrative that gets constructed by the historians and the media.

        The Michael Lewis book was also interesting in documenting when the tide turned – when the brokerage houses suddently realised that they were naked and the tide was going out. There was a sudden deluge of people trying to sell out, to be avoid being left with the can. Consensus changed in a matter of days.

        Taleb’s book also documents how the experts’ predictions arrive at a consensus, and how badly the experts get their predictions wrong. The actual result for any prediction is usually many sigmas away from the consensus (standard deviations measured on the distribution of the predictions). One of the interesting things with economics is that the cycle of prediction to reality is much shorter for climate science, so we are better able evaluate the accuracy of predictions. Climate science is nowhere near, so it is instructive to look at those who predict other complex chaotic systems such as economies to see how well they fare.

  9. Weather Extremes have always happened and they always will. Climate Change has always been with us and it always will.
    The temperature of the earth is, and has been stable in a narrow range for ten thousand years.
    The Black Swan is this: It will most likely stay stable in this same range for thousands of years.
    Analyze the data, there are no instabilities in the temperature data. Modern weather events are well within events of the past ten thousand years. Temperature is well inside the events of the past ten thousand years.
    CO2 is higher, but temperature does not care, it is not driven by CO2

    All you care about is Model Forecasts that prove to be wrong, time and time and time again. Do none of you care about the data?

    If you disagree, use NOAA data and show that temperature is not well inside the range of the past ten thousand years. We have been warming, but this followed a time period when we were cooling.

  10. There is nothing quite like debating a metaphor. The black swan of climate change is scientifically incoherent, so far as I can tell anyway, because speculation is already the problem. Runaway warming, already predicted. Drop into a new ice age, already expected (just not quite yet). No change for a decade, got that already. We don’t need hindsight, because all possibilities are already covered by speculative foresight. The black swan already rules.

    • David,
      That is my point:
      The black swan is that nothing dangerous is going on.
      “The swan that did not trumpet” comes to mind.

  11. The problem with the black swan example for me is that it wasn’t really a catastrophic event. It was just surprising to bird watchers. I think nature has more surprises due to their inherent complexity.

    In computer systems design there is an idea to deal with unexpected events called the ‘robustness principle’ (attributed to John Postel). Which basically says you need to design your system to accept the worst possible and unprobable data/events, otherwise chaos can ensue. Adaptability to change must be a part of the design.

    Sometimes you just muddle through the unexpected the best you can.

    Slightly different is Murphy’s Law – whatever can go wrong will.

    Another systems saying goes – “the probability of failure of a system tends to be proportional to the confidence that its designer has in it reliability?.

    • Teddy –
      Don’t forget O’Toole’s Law, which says that Murphy was an optimist.

    • Please, please don;t keep misquoting Murphy. His first observation was “If someone(e.g. a human) can screw something up, they will.” Mr. Murphy was a test engineer on the Manhattan project and the one who got to climb up the tower after an A bomb test misfired. He found somebody had managed to attach a D-sub connector( a polarized connector like the monitor connector on most PC’s) backwards by hammering it into place. Quite a feat of human engineering.

      The current state of climate science itself is a pretty good example of Murphy’s law in action. Quite a lot of the science has be screwed up by various people, and the rest of us suffer the consequences.

      • >Mr. Murphy was a test engineer on the Manhattan project and the one who got to climb up the tower after an A bomb test misfired. He found somebody had managed to attach a D-sub connector( a polarized connector like the monitor connector on most PC’s) backwards by hammering it into place. Quite a feat of human engineering<

        Is that really true ? Oh god, I hope it is :)

      • Is that really true ?

        Probably not. There are numerous stories about the origin of Murphy’s Law, each almost certainly apocryphal.

        The D-subminiature connector was invented in 1952, seven years after the Manhattan Project completed its work… so this story, while fanciful, is probably no more or less true than the one about how Mr. Murphy was hit from behind while hitchhiking, by a British tourist driving on the wrong side of the road.

      • philc –
        I have lists (yes, multiple) of corollaries to Murphy’s Law. They’re quite extensive and relate to nearly every area of human endeavor.

  12. Black swans may also seem to take place because human memory seems quite poor ; unusual events are forgotten after 5 years or so ; ‘extreme’ events disappear from live memory before the end of a generation (30 years).

    I remember some years ago (1999) when we in France had a strong storm in winter, the famous Leroy Ladurie, a top historian of the middle age climate (!), said : “it’s an unprecedented winter storm !” As a matter of fact, as Emmanuel Garnier, one of his former students, wrote in a later book : it just happened two or three times a century, regularly, during the last 700 hundred years !
    We had the same story last year when we had a tidal flooding on the seashore in Vendée : the same Garnier showed in a report to the French Parliament that this level of coast flooding took place regularly in Vendée during the last centuries ; only problem : it did not happen since the 40s’, and no maintenance work was performed on the coast dykes since that time, due to the perceived absence of risk.
    Lesson to be learned : in many cases, black swans seem ‘unprecedented’ only because nobody remembers what happened more than 15 or 30 years ago. You should only refer to recent floodings in Australia or last week’s tornadoes in the US : the same took place in 1974 !
    So-called experts in climatology do not base their work on actual events from the past ; rather they base their work on scenarios from models !
    If James Marusek’s work on extreme events in the last 2000 events was used as a reference base, it’s likely that no black swan would be identified in the current period.

    • John Kannarr

      And then we have new observational/recording techniques as our technology advances. Events that occurred in the past but were not recorded, at least by the official experts, may now be recorded and assumed to be “unprecedented.” Or our techniques and instruments now produce a more accurate record, say of temperature, than were possible in the past.

    • Daniel –
      Climatology is a young science – really only about 30 years old regardless of what some claim about “climatologists” 50 or more years ago. Therefore, the world (and climate) did not exist prior to 30 years ago. Therefore, all extreme events are “records” and “unprecedented” – regardless of how many times they’ve happened in past centuries. It was William James who said – “The most radical idea in America today is a long memory.”

    • I confirm.
      in Draguignan there has been a drammatic flood, unprecedented… but people remind soon that it hapend at the same place in the 19th century…

      maybe the unprecedented symptom is a consequene of recent evolutions :
      – stronger mediatisation that talk of catastrophes that nobody cares except locals, before…
      – loss of local memory, because people and families move much more over centuries.

      I agree also that many pretended blackwan are in fact caused by a unique and common error. all the pretended improbable failures, beeing simply the predictable consequence of one unanticipated or more often neglected fact.

      unique thinking is the source of many catastrophes.

      however, in theory (the sornete paper explain it), there a case when the system make strong coupling emerge (through chaos) between normally uncorrelated factors… leading in some case to a very unexpected situation, when all fails as one object (because it is one object in fact).

  13. “…Katrina was certainly one, because it hit New Orleans and caused all the damage…”

    LoL….there was a reason their stadium was pre-planned as a shelter in case of such an event.

    I remember reading a article in Popular Science (?) some time prior detailing a “worse case” situation in which New Orleans was to bend over and kiss its a** goodby. The pre-planning staff hit the ball out of the park with its predictions on what a major storm could do to the low laying areas where people had been allowed to build against all good sence.

  14. A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries.

    Support for the above statement:

    I contend that that there is strong evidence for major changes in climate over the Holocene (not Milankovich) that require explanation and that could represent part of the current or future background variability of our climate.
    http://bit.ly/hviRVE

  15. the center of Katrina passed South-east of New Orleans on August 29, 2005

    Found one post several months prior

    http://www.popsci.com/scitech/article/2005-04/hurricanes
    Posted 04.14.2005 at 2:00 am
    It takes Scott Kiser only a split second to name the one city in the U.S., and probably the world, that would sustain the most catastrophic damage from a category-5 hurricane. “New Orleans,” says Kiser, a tropical-cyclone program manager for the National Weather Service. “Because the city is below sea level-with the Mississippi River on one side and Lake Pontchartrain on the other-it is a hydrologic nightmare.” The worst problem, he explains, would be a storm surge, a phenomenon in which high winds stack up huge waves along a hurricane´s leading edge. In New Orleans, a big enough surge would quickly drown the entire city.

  16. This is going to be another run-around-like-chicken-with-their-heads-off thread, it seems.

    Cast not pearls…

    Guys, can we mebbe drop the +/-AGW part of the topic, and consider what policy to discuss relevant to the actual Black Swan subject, rather than point around at everything with a feather or a wing and make meaningless assertions about its chromaticity and lineage?

    Even if Katrina wasn’t a Black Swan to some, the federal government surely did act like it was. A group of people surprised to find a day of the week ending in ‘Y’ if ever there was one.

    Possibly talking about why governments can’t see what any random tourist or science fiction-themed magazine does, or if it sees it doesn’t prepare?

    Twisters have been happening in the midwest long before Dorothy’s trip to Oz; why are hundreds dead without warning in 2011?

    Insurers have been declaiming it too expensive to cover people on coastlines against weather calamity for decades; how is the current state of subsidies on suicidal behavior a solution anyone would call rational?

    Even with a decade or more of updating homeland protection measures (started under Clinton), while I’m not going to go so far as to say that most cities in the USA are still better prepared against a Soviet-era invasion and nuclear attack than commonplace weather emergencies… but I wouldn’t be exaggerating by much to say so.

    Any of you swan wranglers have something relevant?

    • I think the relevance Black Swans, is that in the financial examples that Taleb gives (he’s a quant trader from way back), why is the consensus of predictions so different from reality? Why do experts get it wrong so often?

      Taleb’s answers to dealing with Black Swans are applicable to the world of finance and economics, but doesn’t seem so relevant to climate change.

      Or perhaps we should be looking for potential catastrophies other than CAGW and applying some of the money currently spent on trying to avoid climate change to other more pressing and more predictable tragedies.

      • Experts and modelers are always too certain. Hubris runs amok. Not just in finance, climate science has plenty of examples.

        This is nothing new. Ancient Greek tragedies plowed the same ground. We have a long history of “experts” who predict Malthusian futures only to be wrong. This stems from two trends : 1) experts tend to be bad at predictions (worse than a crowd of common people and about the same as a chimp throwing darts.); 2) Doom and gloomers have a worse record than even the poor showing of experts in general. And of course, the more certain an expert is that he is right, the more likely he will turn out to be wrong.

        It shouldn’t be too hard to see how all this that we know about experts and their predictions has some application to climate science.

    • Bart, as you say: Twisters have been happening in the midwest long before Dorothy’s trip to Oz; why are hundreds dead without warning in 2011?

      Even Dorothy’s folks had a shelter; it was unfortunately Dorothy’s pesky dog what dunnit – kept Dorothy from joining them. Today we learn that the most vulnerable live in trailer parks: are the owners not expected to provide a shelters for the inmates? I’m intrigued.

      I write as one with childhood memories of air raid shelters.

      • Baxter75

        I’ve heard few reports on the tornado casualties, but the reports I’ve heard included five speculative threads:

        1.) Some relied on what were considered stable structures as shelters; the ferocity of the twisters overcame this. The events were more extreme than people expected.

        2.) There was too little warning to get to shelters.

        3.) Some people with shelters were too old or infirm to get to the shelter in time.

        4.) In some cases, shelter was not available.

        5.) Some did not take the warnings seriously.

        I don’t know which of these causes dominated; taking weather seriously and preparing for it in this region isn’t a Black Swan to most; clearly there are opportunities for improvement both in individual responsibility and in weather warning.

        Of course, New Orleans has had almost a century of serious expert warning, and still people there seem perpetually surprised.

        Odd, how even when experts are right, expertise is deprecated and disparaged.

        Wouldn’t the right result be a learning organization?

        A non-government group putting together both experts in science and in policy to take a serious look at weather and climate issues, with a mandate to examine potential calamity and review actual catastrophe, and with a mission to improve and refine the state of knowledge and of policy options by incremental steps?

        We could call it, say, an International Panel on Clim.. oh. Oh, I see there’s a problem here.

      • Yes, this has been discussed a few of blog posts ago.

        You found the problem – The IPCC became an instrument of politics and failed us. As it was devoid of a diverse collection of independently-deciding individuals.

    • a way to resist from unexpected situation
      (call it a blackswan, or like me a dragonking liked to a human mistake ) is to conceive a “system” that is enough resilient, unspecialized, generic, open, adaptable, flexible… so that in case of unexpected it can keep it’s decision power and then keep it’s acting power to adapt to the situation…
      It should also be very diverse (it is very important for ecosystem to be diverse), to be able to survive and react as needed.

      one such a solution is beeing rich (I mean really, this mean having a global population a minimum rich enough to take advantage of modernity… not just a few bilionaires).

      it is the best technic to solve (pretended?) AGW impact in africa… make them rich enough to build their house higher, use better farming technics, have access to school and thus family planning…

      what seems to have missed at fukushima was a generic backup solution… like air-movable diesel, external cooling machines, external commando teams, robots teams…
      what helped was that they could tink the cooling with seawater…
      Thus I’m a bit concerned about sodium cooling, even if modern reactors (like EPR) are much more autonomous and simple in case of trouble…
      every repair should be possible with matches and knifes.

      when you look at weather/climate/nuclear catastrophes, you finally often find that wealth of the community, quality of the government, low war/violence status is the most important factor…

      katrina was a 3rd world catastrophe, and is linked to the social structure of america.
      Xynthia seems to be a similar problem, between local semi-corruption, and state weakening since Reaganomics hits france (dikes were nicely managed for decenies since more than a century).
      Tchernobyl is mainly caused by the encounter of stalinian lack of respect for security, with modern “management by profitability”… after decennies of pure soviet management, the mix of management pressure with soviet terror, cooked with economic weakness lead to the drama… then soviet hiding, occidental cowardness, NGO fear mongering, crony capitalism son of Washington Consensus, finish the dirty job… and most of the dead and poor are not linked to radioactivity, but to crony capitalism and fear mongers.

      what I’ve learn from an old finance engineer, is :
      when things get wrong, all asset are correlated by 1…

      translated in our language , it means : every crisis is a dragon king.

  17. It appears that New Orleans could be devastated by a foreseeable Black Swan — the diversion of most of the Mississippi flow to the Atchafalaya River.

    http://www.tulane.edu/~bfleury/envirobio/enviroweb/FloodControl.htm

    It may even happen this year.

    • John Kannarr

      I’ll just put in a pitch for the highly readable book, The Control of Nature, by John McPhee (and one of the references in the above article). Its chapter on Atchafalaya tells the story of the possibility of “stream piracy” by the Atchafalaya River in diverting the lower Mississippi River, completely bypassing New Orleans and its ports.

    • A bad outcome is not a Black Swan. The Mississippi diverting into the Atchafalaya is foreseeable and inevitable, not improbable.

      Geologically it has happened many times in the past 10 million years. It has to happen again. My Geology Prof in 1974 spoke of the Army Corps of Eng. fighting to keep the Mississippi out of the Atchafalaya. “They may delay it, but they can’t stop it.”

  18. What do you do with a black swan once you decide you found one? These occur in project management as well. A pragmatic approach is list them associated with what you expect to occur. If there is little or no cost to adjusting things to be able to adapt more effectively, then adjust the plan. There are many possibilities which are foreseeable, but they usually aren’t foreseen unless you try.

    Many things do not change in a smooth fashion. People don’t grow taller linearly over time – there are growth spurts. Knowledge grows in spurts associated with paradigm shifts. Atlanta rain just starts dumping all of a sudden. These are the features which are most interesting, but averages and models don’t really deal with them well – they work best for the uninteresting features.

  19. “what may be a Black Swan surprise for a turkey is not a Black Swan surprise to its butcher”

    I’m assuming the IPCC is the cleaver. Now if I could just figure out who the turkeys are…

    • Harold –
      I’m assuming the IPCC is the cleaver. Now if I could just figure out who the turkeys are…

      The purpose of the exercise here is to make the “skeptics” the cleaver – and the IPCC the turkey.

      Note that for the IPCC, CRU, etc – Climategate was a Black Swan.

  20. In our life time, there is a small chance that the Yellowstone Caldera will erupt and, in so doing, extinguish the human race. This is a black swan.

    The more interesting question is ‘what should we do about Yellowstone?’ This question is illuminated when one considers that there are an infinite number of potential black swan events (by definition) that we potentially could ‘do something’ about. Attempts to remediate any of several such black swans could consume the entire wealth of nations and still not make a whit of difference (except to make some feel better because they are ‘doing something’, however misguided).

    Having read the book, I found that Taleb’s notion of black swans asymptotically approaches worthlessness the more you think about it.

    So to answer the question ‘what should we do about the potential for an eruption of the Yellowstone Caldera’, my answer is ‘absolutely nothing’.

    • I disagree. Yellowstone is survivable for some. Just not those within a few hundred miles (or 1000).

      We don’t spend money on solar panels – the ash cloud from Yellowstone will render them useless if people do survive the explosion.

      We stop squandering trillions on global warming and get a colony in space going. Maybe mankind survives.

      We spend money on detecting killer asteroids (and the ability to destroy or deflect them), not on wind turbines.

      Remember, this interglacial will end. It has to. It will. Not too long from now.

      • In terms of climate, I think major volcanic eruptions are most definitely Black Swans.

        When “the team” makes a bet, they almost always include a caveat for a major volcanic eruption.

  21. Pielke Jr has a relevant post (relevant also to the catastrophe modeling thread), this excerpt from the Financial Times:
    http://rogerpielkejr.blogspot.com/2011/05/financial-times-on-disasters-and.html

    “The biggest question, though, remains the extent to which climate change is the driver of hurricanes, cyclones and flooding that have hit the world with apparently increased ferocity and regularity in recent years.

    It is still proving extremely difficult for scientists to extract a clear sign of the effects of climate change from the normal long-term historic cycles of weather and climate activity. That is despite simple logic saying that a warmer climate should result in more powerful storms because of a greater water content in the atmosphere.

    Axel Lehmann, chief risk officer at Zurich Financial Services, says it is necessary to take a long-term perspective – of 200 or even 1,000 years.

    “In terms of severity and frequency, is this type of event happening in a more systematic way? We do not yet have an answer on that,” he says.

    “But on a systematic basis we do know that a growing population puts pressure on the earth and its resources.”

    • Dr. Curry,
      Why is that quote from the Financial Times accepted as a starting point when it is false?
      There is no record to show with facts that “hurricanes, cyclones and flooding that have hit the world with apparently increased ferocity and regularity in recent years.”
      If that false premise is allowed to stand unchallenged, then the rest of the conversation is useless.
      Mr. Lehmann is frankly evasive in his answer. There is no trend in storm strength severity or frequency, but he does not wish to admit it.
      Until this false assumption is challenged for the counter-factual premise it is, we are not going to advance this discussion.

      • That wasn’t so much false as it was a tapdance. The media can be pretty good at tapdancing; implying one thing while technically saying another. Gavin Schmidt is the Fred Astaire of the verbal tapdance.

  22. The Black Swan discussion is interesting but its lessons aren’t simple.

    The power of the Black Swan example is that for people who thought that “White” was part of the definition of swans, a Black swan was truly unimaginable. But this simply reflects a folk theory of species. I don’t know what year the black swan was found, but presumably at that time zoologists were adequately clear on the definition of species to know that color was not essential. If you asked an ornithologist “could a swan be black and still be a swan”, the answer would surely be yes. The question “how likely is it we’ll find one?” is a different question. But, you know, think about albinism and things that shouldn’t be white and are.

    Given the deep knowledge acquired by scientific inquiry, mostly in the last 100 years, there are not likely to be many more “true” black swans, where literally nobody had the tools to “imagine” the event. It might be an interesting exercise to see what that’s happened in the last 50 or 100 years was actually a black swan in that sense. My belief is that it will be a very short list. Certainly market crashes, cold winters, Katrina, this month’s tornadoes, are not black swans in the original sense.

    If you want to use the second meaning of black swan as “something the experts thought was so unlikely as to be ignorable”, you get a much larger category. Even then, you end up with debates like “how many experts have to predict something for it not to be a black swan?” which aren’t really all that constructive.

    Obviously the claim made by so many here that “no AGW would be a black swan” is dealing in this second domain. Obviously there are plenty of “skeptics” (I still don’t know the politically correct term for those who think AGW is very unlikely) who are predicting no AGW, so in the big sense it would no more be a black swan than Katrina. What I think is lost on those skeptics who read the IPCC selectively is that no AGW would not be a black swan to most climate scientists: to put it in reference to probability distributions for the climate sensitivity, every serious climate scientist has a probability distribution with some density at 0. It may be small, but I – and presumably the vast majority of relevant experts – can certainly “imagine” circumstances in which (say) doubling CO2 does not lead to significant warming.

    My question then, for all those who are so sure that there will be no CAGW: if there were, would it be a black swan for you?

    • I think climate zombies is the politically correct term that you are looking for.

    • John Kannarr

      I suspect that for many, if not most, skeptics, like myself, it isn’t the case that we can’t imagine AGW, or even CAGW, occurring. It is just that we haven’t seen any convincing evidence for such, but have seen lots of what we consider to be attempts to fudge the data, overwhelm logic with sensational scare stories (aided by a complicit MSM and environmentalists in general), invalid mathematical calculations, and flawed claims that models are sufficiently dependable to provide any useful information about the likely future.

      So, no, neither AGW or CAGW would be a black swan for us … if we were ever presented with a recognizably unbiased case for them. We can conceive of them, but certainly aren’t willing to accept them based on what we’ve seen so far.

      When advocates for AGW start following Feynman’s prescription for scientific integrity, for example, then we’ll see what the totality of the data and arguments add up to.

    • Paul –
      A couple points –
      Obviously there are plenty of “skeptics” (I still don’t know the politically correct term for those who think AGW is very unlikely) who are predicting no AGW

      The word “skeptic” will work well although 1) many of us apparently don’t do “politically correct” and 2) I think you’re conflating two different groups of skeptics in that one phrase and 3) GW is “generally” accepted by most skeptics and some A contribution is obvious to many of us, but it’s the extent of the A as the cause that’s in question.

      What I think is lost on those skeptics who read the IPCC selectively

      Some skeptics have never read the IPCC. But those who do/have/will generally don’t read it selectively. At least not that I’ve noticed.

    • John Kannarr

      I suspect that it’s not the case that most skeptics of AGW or CAGW, like myself, do not think that there could be such. It is that we haven’t seen convincing evidence or arguments that don’t appear to have been contrived, fudged, based on invalid calculation methods, or based on models (or proxies) that haven’t been validated, by people who haven’t owned up to past errors in prediction but are apparently continually rewriting history so that the latest weather calamity is suddenly discovered to have been predicted all along. If AGW or CAGW were to occur, then, it wouldn’t be a black swan at all, though it might be a big surprise to us.

      I realize that a lot of the sensationalism that appears in the MSM is due to the MSM itself or to environmentalist advocates in general, but somehow I never see the climate scientists rushing to disabuse the MSM of the latest wild claim and to tell the public that the sensationalism has no reasonable basis in the science. There seems to be a bit of plausible deniability going on, as though the hope is that the public will be riled up to support “action” but that climate scientists can always say later if called on failed sensationalist predictions, “well, we didn’t actually say that in our refereed journal papers.”

      Most of us skeptics are still waiting to see climate scientists begin to live up to the Feynman definition of scientific integrity and to provide complete transparency about data and calculation methods, as well as to respond to the skeptical scientists in the journals.

    • The black swan would be that 100 years from now the IPCC assessments should turn out to be anything other than stark lessons in how not to do science.

    • Paul Baer

      Given the deep knowledge acquired by scientific inquiry, mostly in the last 100 years, there are not likely to be many more “true” black swans, where literally nobody had the tools to “imagine” the event.

      What?

      Read Taleb’s book, Paul, and you will see that there are very likely to be major “black swans”, especially as they would apply to the embryonic climate science today.

      Max

      Max

      • I did read most of the book a couple of years ago, though I confess to not remembering many of the details. But despite what you call the “embryonic” state of climate science, between the laws of thermodynamics, fluid mechanics, and atmospheric chemistry, the basic dimensions of the climate system are pretty well understood. That there will be unpredicted state changes at a variety of scales is inevitable. But it seems unlikely to me that there will be many events to which someone says “we never thought THAT could happen!”

      • You should re-read the book because that was the entire point of it.

        You sound very conservative. Life is full of surprises.

  23. When you live in a bubble anything that pops it is a black swan. For those outside the bubble it’s like watching the inevitable car crash, you yell stop, stop, but bubble people by the nature of their cage can not hear.

    There are actually very few black swans if any at all, only conventional wisdom which is not very wise at all, more often than not.

    • Jerry,
      Well said.

    • Very true. If you’re an alcoholic, you don’t understand why you’re having problems. If you’re not, it’s plainly obvious.When he cause of a black swan is retroactively obvious, it’s often obvious before the fact to the same group of people to whom it’s obvious after the fact.

      Example: it was obvious to me that the stimulus would cause inflation before they enacted it. It was equally obvious that it wouldn’t stimulate anything but inflation. And these things were obvious to lots of people. Now that inflation is knocking at the door with 8+% unemployment, the Krugmans of the world are baffled.To them it’s a black swan, and a completely unexplained phenomenon. To the rest of us, it was obvious.

  24. O dear I am confused.

    The black swan problem refers to the existence of the unknown that may always surprise. Such as the discovery of black swans on the Swan River in south western Australia in the 19th Century. I am sure that aboriginals had discovered that they were very tasty indeed some 50,000 years previously – but we will let that pass.

    Unknowns in climate are by definition unknown. But there are some examples of recent discoveries that are still working their way through the zeitgeist. These include top down climate forcing by solar UV, dynamical complexity in climate, decadal, centennial and millennial variability of Earth systems, a little current around Cape Agulhas that Max introduced my to yesterday and cold Southern Ocean water accumulating in the region of the Humboldt Current in the richest fishery in the world. There seems little doubt that there are many discoveries to be made and some of these will be most surprising.

    The Black Swan Theory is however a usurper and interloper. The issue of knowledge and Earth system extremes is being falsely conflated – it is a two different problems. One is for the unknown – which is a vast domain. The other is for extreme variance often associated with chaotic bifurcation that Sornette called Dragon Kings. Conflating the two is a recipe for a confused and confusing narrative.

    The floods in Queensland were, for instance, neither unpredicted nor extreme. Australia is a land of extremes – so much so that the extreme is normal. North east Queensland rainfall falls into a couple of distinct patterns. One of these is a decadal pattern where the average rainfall in dry periods is a quarter of the average during wet periods. This has been understood since the 1980’s and is something that has frequently been communicated by hydrologists. That this systematically derived knowledge has been drowned out is not a problem of black swans but of the way science has been commandeered by the climate warriors.

    The rainfall variability emerges out of sea surface temperature variability in the Pacific Ocean. Different states punctuated by extreme ENSO variability in 1976/1977 and 1998/2001. A warmer and drier state to 1998 and cooler and wetter thereafter. There were extreme states – or Dragon Kings – in 1976/1977 and 1998/2001 followed by a different climate trajectory. We are likely in for a decade or 3 of these ‘not extreme’ conditions. The changes were abrupt and then Earth systems settled into a new state – the definition of complex dynamical systems.

    These are not entirely abstract ideas for me – they erode their way into the morphology of the Australian landscape. They have an expression in pools and riffles or as braided high energy streams. Something I find very compelling. That there seems to be a global pattern of variability in ocean and atmospheric indices – and biologies – with a similar temporal signature suggests the existence of a black swan. I am currently favouring top down Earth systems forcing by solar UV.

    • maksimovich

      The poultry problems are well described in the mathmatical literature of mixed mode oscillations in slow fast systems .
      eg J. Guckenheimer, Yu. S. Ilyashenko, The Duck and the Devil: Canards
      on the Staircase, Moscow Math. J., Volume 1, Number 1, 2001

      That they are generic families (related) is also well described eg E. A. Shchepakina, V. A. Sobolev, Standard chase on black swans and
      canards.

      In this paper we use the standard approach to study slow integral surfaces of variable stability (or black swans). These surfaces are considered as natural generalizations of the notion of a canard.

      We suggest to use the term ”black swan” by two reasons. The first one is
      that a swan is a bird of the family of ducks. The second one is connected
      with the usual meaning of ”black swan” in the sense of a rare phenomenon.
      It should be noted that the French term ”canard” is used in the sense of false
      rumour in English.

      http://www.scholarpedia.org/article/Canards

      • ‘We develop the concept of “dragon-kings” corresponding to meaningful outliers, which are found to coexist with power laws in the distributions of event sizes under a broad range of conditions in a large variety of systems. These dragon-kings reveal the existence of mechanisms of self-organization that are not apparent otherwise from the distribution of their smaller siblings… We emphasize the importance of understanding dragon-kings as being often associated with a neighborhood of what can be called equivalently a phase transition, a bifurcation, a catastrophe (in the sense of René Thom), or a tipping point. The presence of a phase transition is crucial to learn how to diagnose in advance the symptoms associated with a coming dragon-king.’

        http://arxiv.org/ftp/arxiv/papers/0907/0907.4290.pdf

        Now we have ducks – the confusion is deepening. (Now is not the time to be silent Bart). I suggest that the term ‘rare’ is the operative word here. It is a bit misleading because blacks swans are neither rare not endangered – and they only escaped attention for so long because they live in and around the most goegraphically isolated city in the world. Does it refer to the black swan problem or to the theory of black swan events? The black swan problem relates to the existence of black swans. The proposition that all swans are white is falsified by a black swan. It has come to be a metaphor for the unknown unknowns in science. ‘The Black Swan Theory or Theory of Black Swan Events is a metaphor that encapsulates the concept that the event is a surprise (to the observer) and has a major impact. After the fact, the event is rationalized by hindsight.’

        Dragon Kings on the other hand are the extreme variability that occurs at phase transitions in complex dynamical systems. Now I am sure the duck was complex – but it seems more a description of the topology of the phase space than the variability of the system at the phase transition.

        The 1997/1998 El Nino is an example of a Dragon King. Duck a l’Orange is an example of a canard. Dragon Kings are a better description of extreme climate events because everything in weather and climate is chaotic. Dragon Kings are a property of complex dynamical systems signifying extreme and rapid changes that occur at phase transitions. They are in principle predictable from first principals. This is quite different from the either sense of the use of the term black swan.

        Nothing in climate can be categorised as a black swan – everthing is in fact predictable. I predict that we will have extreme everything.

      • maksimovich

        In September 2009 the UK Met Office, expecting a positive NAO to be dominate, predicted winter temperatures to be near to above average over much of Europe.

        The sign was 180 degrees out of phase is well known.With a significant velocity inversion .
        http://www.cpc.noaa.gov/products/precip/CWlink/daily_ao_index/ao_index.html

        The subsequent december 2010 T inversion in the uk of around-5.9c
        suggests that these phase inversions can repeat.

        Now if persitence became a normative state, where a number of albeit rare externalities became phase locked in an inverse mode interesting things can happen,

        http://onlinelibrary.wiley.com/doi/10.1002/joc.1788/abstract

      • maksimovich

        It seems there is another way to explain the 2009 problem.

        http://www.metoffice.gov.uk/science/creating/monthsahead/seasonal/2009/winter.html

      • Are cold winters in Europe associated with low solar activity?
        M Lockwood, R G Harrison, T Woollings and S K Solanki

        http://iopscience.iop.org/1748-9326/5/2/024001/fulltext

        I think this might be a black swan in climate science. There is a Southern Hemisphere equivalent that interacts with ENSO.

      • Interestingly, speaking of ducks, there are penrose tiles that look a lot like chickens.
        http://beehive.thisisgrimsby.co.uk/default.asp?WCI=SiteHome&ID=14894&PageID=94482
        Penrose is also famous for his critique of another popular science consensus, that of AI.
        http://www.amazon.com/Emperors-New-Mind-Roger-Penrose/dp/0140145346

      • If it walks like a duck…

      • So, would that be a UEA swan, a Penn State swan, or a Guelph or Toronto swan? Finger test?

      • maksimovich

        In addition to “black swans”, “ducks” (of all colors), “canards” (of all sorts), “turkeys” (both within and outside the Taleb definition), “chicken littles” (or, in Britain, “henny pennys”) you also have “quacks”

        quack (kwak)
        1. one who misrepresents their ability and experience in diagnosis and treatment of disease or effects to be achieved by their treatment.
        2. a charlatan

        We live in a fowl society.

        Max

      • But does fowl weather cause fowl climate?

    • >Different states punctuated by extreme ENSO variability in 1976/1977 and 1998/2001. A warmer and drier state to 1998 and cooler and wetter thereafter. There were extreme states – or Dragon Kings – in 1976/1977 and 1998/2001 followed by a different climate trajectory<

      Almost cyclic tipping points, one might venture to say :)

      • These shifts in climate are well known in the literature of Australian hydrology – and to farmers too I might add. But they are not cyclic – they are in something known as chaotic bifurcation. This is an adrupt and non-linear change in climate as a result of sensitive dependence in chaotic spatio-temporal Earth systems.

        What this means is that small changes magnify through interacting systems to produce large changes in the system. Typically you get a large pertubation after which things settle into a new pattern.

        This is why there is a risk from greenhouse gases. There is a risk – that I can’t quantify – of places in the world plunging 10’s of degrees in temperature in a matter of months. I stress that this is just one possible outcome and that things are very unpredictable.

      • Thanks for these comments – I have known of these shifts for a long time (my earlier post was sardonic)

        I still believe that these shifts have a cyclic component, despite the circumlocutious term “chaotic bifurcation”, because they keep happening. That you have characterised their cause as magnification of small changes by system interactions doesn’t rebut this view – time scale is a very slippery customer and finding the most accurate time fit for cycles of complex interactions is likely to take much longer than 30 years (the preferred time grab so far for AGW modeling)

        There are many places in the world where temperatures already plunge 10’s of degrees in a matter of months. Most Aussies do not grasp that, eg. in Siberia (amongst other places), the high August daytime temperatures of 30+C give way to -15/20C in October. If you’re working in this climate, it can be a real nuisance, but scary it isn’t

        I appreciate the comments you make here. I read your “angels on a pinhead” posts in the Brain Sprain thread and wondered if non-rhetorical discussion was possible

      • I used the words chaotic bifurcation as a description of a state in the theoretical physics of complex dynamical system.

        If they are just words to you – there is a thread here on spatio-temporal chaos. Without these concepts understanding anything about climate is impossible. Climate simply works in a different way than what most people imagine. Instead of simple causality we have abrupt and non-linear climate shifts.

        https://judithcurry.com/2011/02/10/spatio-temporal-chaos/

        The Younger Dryas is named after a plant found only in alpine or tundra regions which was found in bog layers from the era in Europe. There is a bog in Ireland showing average temps falling dramatically in a matter of months. Average temps in Ireland like winter in Siberia would be a problem for the Irish.

        There are lots of sources on the web – google abrupt climate change – here are a couple one for paleoclimatic and one for modern ocean/atmosphere coupling.

        http://books.nap.edu/openbook.php?record_id=10136&page=R1
        http://www.pnas.org/content/97/4/1355.full.pdf

        If you want to insist that you understand and can predict climate into the future – you are nothing but a misguided warrior in the climate wars speaking in a superficial idiom of science. As Voltaire says – doubt is uncomfortable but certainty is absurd. Doubt I would add is the mark of a seeker of truth.

      • These bifurcations have what is called phase space – they are like the butterfly wings of the Lorenz attractors – http://en.wikipedia.org/wiki/File:Lorenz_attractor_yb.svg – which has just 2 wings or ‘strange attractors’. The phase space is the overall topology of the solution space – the 2 wings.

        Climate has many ‘strange attractors’ – they define the mode into which climate may ‘shift’ as a result of something called sensitive dependence. These are the small changes I mentioned.

        If you think this is gobbledegook – well you see the 1st problem in climate science.

      • Part of my initial quote:

        “… wondered if non-rhetorical discussion was possible”

        Your quote (1):

        “Average temps in Ireland like winter in Siberia would be a problem for the Irish”
        I didn’t compare Ireland to Siberia; that you have used a straw man to try and rebut me is quite disappointing. To change Ireland’s climate to Siberia’s climate would take quite a bit longer than a few months of tectonic drift

        Your quote (2):

        “If they are just words to you – there is a thread here on spatio-temporal chaos”
        I’ve read it; that you assume I haven’t is disappointing. Yes, I understood it – if you have read some of my posts on the chaotic but ruthlessly opportunistic nature of the evolutionary process, you would have realised that I have understood it

        Your quote (3):

        “If you want to insist that you understand and can predict climate into the future – you are nothing but a misguided warrior in the climate wars speaking in a superficial idiom of science”
        I have insisted no such thing (nor do I believe it), but I have pointed out that I think past time slices offer patterns that may be of use. That you insist they are not useful for future guesstimates causes me to wonder at the direction and placement of misguidance. And I am no climate warrior – rather I have an atavistic dislike of those who would profoundly disrupt my childrens’ livelihoods without regard for intellectual honesty or practical alternatives to very damaging drops in living standards

        Your quote (4):

        “If you think this is gobbledegook (strange attractors)– well you see the 1st problem in climate science”

        I don’t use the word gobbledegook for this language (I reserve it for marketing or bureaucratic attempts at evading reality) but the issue with your point here becomes this:

        We cannot be sure of the result of continuously adding CO2 to the atmosphere over time, so we should stop doing this … but we cannot be sure of the results if we do stop, since prediction is impossible and other “strange attractors” will likely bite us on the bum anyway

        That’s why I labelled it circumlocutious. Perhaps pointless is a better description ?

        Your unhappy but liberal use of straw men has made me wonder if the the answer to my initial question is NO!! At this stage, I’m still inclined to think that this would be a pity, but perhaps you may care to change my mind on this

      • The word used was idiom not idiot – :roll:

        The 1st problem of climate sceince is that no one understands it. The second is that everyone petends they do.

        The idiom and the superficiality reference comes from a paper called “The Wrong Trousers’ – www2.lse.ac.uk/researchAndExpertise/…/mackinder_Wrong%20Trousers.pdf This is a predecessor to the Hartwell 2010 paper.

        ‘Although it has failed to produce its intended impact nevertheless the Kyoto Protocol has performed an important role. That role has been allegorical. Kyoto has permitted different groups to tell different stories about themselves to themselves and to others, often in superficially scientific language. But, as we are increasingly coming to understand, it is often not questions about science that are at stake in these discussions. The culturally potent idiom of the dispassionate scientific narrative is being employed to fight culture wars over competing social and ethical values. Nor is that to be seen as a defect. Of course choices between competing values are not made by relying upon scientific knowledge alone. What is wrong is to pretend that they are.’

        My first principle is that we are performing an atmospheric experiment for which we have not the wit to understand. You equate chaos to evolution? The Younger Dryas to continental drift? You understand nothing and arrogantly assume that you know everything – that is the problem with the climate wars.

        That there are alternatives to the widespread destruction of capital is abundantly clear. I don’t intend to canvass it here – read the Hartwell 2010 paper. If I thought it was possible to change your mind – I might try – but I am assuming it is set in concrete.

      • I’m sure there’s room somewhere for the description “superficial idiot of science.” But perhaps not in a relatively polite blog like this one.

      • So I’ll politely ignore the pointless “idiot” epithet

        But you are invited to demonstrate the “superficial” adjective.

      • It’s only a play on words which amused me, not a response or reference to anyone or anything on this blog. Sorry if that wasn’t clear, it certainly wasn’t aimed at you.

      • I also meant to add that my posts here will be somewhat intermittent over the next 10 days or so as I’m working in someone else’s quite distant backyard, but a period of non-posts doesn’t mean I’ve run away

        So again, demonstrate the “superficiality”, please

  25. Fundamentally, doesn’t it boil down to
    the known unknowns (vague, incomplete, unquantified aspects – clouds come to mind) and
    the unknown unknowns – perhaps this is where the real ‘black swans’ reside?

    It seems to me that neither of these can be built into reliable models and if the former are deemed relevant and significant then the modeling endeavour is a waste of time at best and misleading or fraudulent at worst.

  26. After the fact, the event is rationalized by hindsight.

    Global warming causes increased snow: http://bbc.in/iNhvIx

    • Girma

      Global warming causes increased snow:

      So is this negative feedback built into the IPCC models?

      • gyptis444

        How can you ask such a question?

        Don’t you know modern science also involves once wishes?

  27. Harold Pierce Jr

    Black Swan Event: Canucks winning the Cup.

    The Canucks have never got past Round 2 of the play-offs.

  28. When judging what is out there that could be completely unexpected by the experts- which is, I think a good definition of the black swan concept in this concept, it might be good to look at what others have pointed out in the past regarding doom and calamity.
    A good place to start would be to read someone who was well positioned to have a good perspective of how experts across a broad spectrum thought.
    Say, an editor of a prestigious and respected science magazine like “Nature”.
    John Maddox comes to mind as just that sort of observer.
    http://en.wikipedia.org/wiki/John_Maddox
    Accomplished, educated, published and for 22 years Editor of Nature magazine.
    Now he did write a book about unexpected things.
    That book is “The Doomsday Syndrome”.
    http://www.wired.com/wired/archive/11.07/doomsday_pr.html
    It is available still
    http://www.amazon.com/Doomsday-Syndrome-John-Royden-Maddox/dp/0070394288
    He points out that scientists seem to periodically rally around some new great danger that is going to kill us all, and that these rallies have similar characteristics.
    I would suggest that one way to look at this is to see that the ‘black swan’- the thing that exists but that experts do not see- is that we are going to do OK with this climate, that a cliamte doom is no more likely than any of the other failed doomsday prophecies that litter history.

  29. Global warming obviously is not a black swan. It is an event “outside the realm of regular expectations” but one can’t say “nothing in the past can convincingly point to its possibility”

    Global warming is NOT outside the realm of regular expectation.

    Just look at the following graph for a minute.
    http://bit.ly/iUqG8I

  30. It seems any event can be a “Black Swan” because any counter-arguments can be dismissed a post-hoc rationalization.

    The 2011 SEQ floods are a supposedly “outside the realm of regular expectations, because nothing in the past can convincingly point to its possibility”. SEQ has floods almost every year – and there had been major floods in May 2009 – not as severe but sufficient for state of emergency to be declared, and more than sufficient you’d think to suggest more extreme event might be possible.

    Months before the 2011 floods the Qld Govt had published a report indicating an expectation that climate change would make extreme rainfall events more frequent in SE Qld – yet apparently the floods show “Our expectations of extreme events in a warmer world can be soundly trumped by natural variability of weather processes.”

    Where is Alan Sokal?

  31. All a bit of nonsense. ‘Black swan’ idea is based on a subjective perception of one or a group of individuals. Earth Sciences should objective and devoid of ‘prejudice’ rooted in our contemporary lack of knowledge or understanding. Perhaps ‘Dark Ages’ were so dark due to countless ‘black swans’ to be found all over the place.

    • Quite so. This thread is very interesting and sometimes amusing, but I’m not sure that dwelling on Black Swans, real or imagined, will clarify the AGW debate.

      Perhaps paradoxically, the first (literal) black swan I encountered was in the severe English winter of 1962-63, when Christmas snow in London remained on the ground for months, and the lake in my local park where the aforesaid swan resided remained frozen until May. Not something a Perthian aboriginal could have foreseen.

    • V: You are correct that which swans are black (as it were) depends on what one believes is likely, or unlikely, because inductive probabilities are subjective. The fact that climate science is presently polarized makes this an unsettled question. But it is far from nonsense. It is how things are.

      • Hmm. Just because this is the way ‘things are’ doesn’t mean it’s not nonesense.

        V makes an excellent point.

  32. I would nominate a couple of major known physical events as potential Black Swan events of high, even inevitable probability but occurring sometime over periods that cannot be estimated or currently foreseen.

    Solar mass ejections [ SME’s ] of similar massive power to the 1859 Carrington event.
    SME’s of very high power appear to be more common during low solar activity cycles.
    An SME of the Carrington event’s power today would in all likelihood destroy most of the vital electrical power systems and electronic communications networks across most of the SME’s target hemisphere. That would lead to a hemispheric wide breakdown in the affected areas infrastructure, governmental control and it’s ability to direct recovery and total failures in the modern electrical / communication dependent food supply and food transport and water distribution networks leaving billions without the sophisticated infrastructure that underlies all of today’s city centric society ability’s to exist and survive.

    Super volcano Laki in Iceland to again resume it’s activity of 1783 / 1784 with consequent crop failures across the whole of the northern hemisphere and mass starvation of a fair proportion of the global population as a consequence.
    And that is far, far more likely than Yellowstone erupting any time in the next few hundred years.

    Even midway through solar cycle 23, the forecasts for solar cycle 24 were that it would be one of the strongest in the last thousand years. It’s not and is now down to the very low activity levels of the Maunder and Dalton minimum cycles.
    A Black Swan event in solar physics and a what may eventually be seen as a totally unforeseen Black Swan event in the evolution of climate science?

  33. Kim does these better than I can, but here is a go:

    The climatologist quacked
    The black swan pondered
    Was that a canard?

    Max

  34. Judith Curry

    [Tried posting this earlier, but it got hung up in the spam filter]

    Your summary, “Anticipating the Climate Black Swan” brings up some key issues in the ongoing scientific and policy debate surrounding the premise of alarming anthropogenic greenhouse warming as postulated by IPCC.

    You wrote:

    Overall, the idea of climate black swans hasn’t been explored very much, but I think it is something that deserves further consideration. I agree that global warming itself shouldn’t be regarded as a black swan, although if some of the more alarming sensitivity predictions were to be true with accompanying extreme weather events, then it might arguably be a black swan.

    I think global warming, per se, is probably not a “black swan”, as you say, Judith (because it is happening today, for whatever root causes), but our planet’s climate may well be so unpredictable (based on the very limited knowledge we have today) that it is pre-programmed to introduce many “black swans”.

    The FT summary of “black swan” wisdoms you have cited is interesting, but it relates to economic issues, rather than specifically to our climate.

    But let’s get to Taleb, himself, and the black swan concept as it applies to climate science today.

    Taleb emphasizes that the world is dominated by the extreme [“Extremistan” as he calls it] despite “the ingrained tendency in humans to underestimate outliers – or Black Swans” and advises us: “Black Swans being unpredictable, we need to adjust to their existence (rather than naïvely try to predict them).”

    The most important “take homes” for me from Taleb’s book, which relate directly to the role of the IPCC in making alarming predictions related to global warming (to help policymakers come up with mitigating policies):

    “we are demonstrably arrogant about what we think we know” (Ouch!) (Ch. 10)

    “There is no difference between my guessing a variable that is not random, but for which my information is partial or deficient…and predicting a random one.” (Ch. 10)

    “information is bad for knowledge” or “the expert problem”: i.e. “ additional knowledge of the minutiae of daily business can be useless, even actually toxic” (in making predictions); the “expert” is no better at making predictions than anyone else. (Ch. 10)

    “What matters is not how often you are right, but how large your cumulative errors are.” and “The errors get worse with the degree of remoteness to the event.” (Explains why longer-term predictions will be more wrong than shorter-term ones) (Ch.10)

    “You cannot ignore self-delusion. The problem with experts is that they do not know what they do not know.” (This statement is so simple and self-explanatory, but it describes in a nutshell the problem with the IPCC “experts” and the “uncertainty” problem you have emphasized). (Ch. 10)

    “Prediction requires knowledge about technologies that will be discovered in the future. But that very knowledge would almost automatically allow us to start developing those technologies right away. Ergo, we do not know what we do not know”. So much for the IPCC “scenarios” of atmospheric CO2 levels by year 2100 (upon which the long-range temperature forecasts are based)!(Ch.11)

    “be a fool in the right places”: i.e. it’s OK to be a fool by making a failed prediction, but ”what you should avoid is unnecessary dependence on large-scale harmful predictions” (Self-explanatory for the IPCC role in climate science) (Ch. 12)

    Taleb distinguishes between what he calls “experts who tend to be experts” and “experts who tend to be…not experts”, drawing the difference between “things that move” (where there are no real experts) and “things that do not move”. Although he does not get into this topic, it is clear that “climate moves” and that the climatologists of today fall into the latter category.

    He cites the “I was ‘almost’ right” or ”other than that, it was okay” syndrome used by predictors when their projection turns out to be wrong. As just one example, the failure of the IPCC prediction of 0.2C warming for the first decade of the 21st century has been rationalized in many ways: “it was correct except for… unplanned natural variability, an unexpected shift in ENSO, above-normal human aerosols, etc.” (add in any rationalization that sounds good at the time).

    Common ploys used here are: “you invoke the outlier” (see above) or “you tell yourself you were playing a different game”: “well this was just a model projection, not a prediction”.

    Taleb gets into discussion of statistics, pointing out some common misconceptions about the bell curve and randomness in real life as well as difficulties in parameterization in models. As far as the role of models (not specifically to climate science) he writes: ”Most models, of course, attempt to be precisely predictive, not just descriptive; I find this infuriating. They are nice tools for illustrating the genesis of Extremistan, but I insist that the ‘generator’ of reality does not appear to obey them closely enough to make them helpful in precise forecasting.”

    Taleb points out: “Today academics in abstract disciplines depend on one another’s opinion, without external checks, with the severe occasional pathological result of turning their pursuits into insular prowess-showing contests. (Have there been examples of this in climate science?)

    In describing his own tendency toward “skeptical empiricism”, Taleb writes: ”A theory is like medicine (or government): often useless, sometimes necessary, always self-serving, and on occasion lethal. So it needs to be used with care, moderation, and close adult supervision.”

    Taleb’s “The Black Swan” is brilliant. I have re-read it several times. He has not written about climate science, per se – in fact, I do not believe he even mentions it, or (if so) it is just in passing.

    But what he has written applies across many fields, including climate science today.

    You are spot on by bringing this concept into the climate discussion.

    You mention an editor who responded to your:

    A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries.

    with this comment:

    Good point, you should elaborate how this would help science, policy

    Help science: We should admit that there are great uncertainties in what makes our climate behave the way it does, particularly in the relative importance of natural versus anthropogenic forcing factors and potential feedbacks. We should attempt to better identify these unknowns in the context of observed longer-term climate change of the past, measured in centuries. But, at the same time, we should concede that until these basic uncertainties are clarified more closely, which may never be possible because of unknown and hence unquantifiable black swans, we have no meaningful knowledge about how human CO2 emissions could affect our climate in the future.

    Help policy: We should acknowledge to policymakers and the general public alike that climate science is still in its infancy today and that the unknowns outweigh the knowns by several orders of magnitude. All we know for sure today is that we do not know. Until more is known, it would be foolish to make any projections of future climate and even more absurd to base any potentially costly long-term policy decisions on climate science today.

    To quote Taleb: develop and use climate projections “with care, moderation, and close adult supervision”.

    And I would add: “Keep them out of the hands of “policy-makers”.

    Just my thoughts, Judith.

    Max

    • Max – the excerpting of quotes is very helpful, thanks for that.

      I’m with you of course right until “we have no meaningful knowledge about how human CO2 emissions could affect our climate in the future.”.

      The basics of the greenhouse effect are meaningfully known. More greenhouse gases, more energy trapped. Probably the surface temperature will rise, the hydrological cycle will speed up. Of course all kinds of other effects will take place, and as I said elsewhere on this thread “no AGW in response to doubling of CO2” is within the subjective probability estimates of climate scientists. But the mean of the distribution for most is in the realm of “we should probably stop increasing GHG concentrations”.

      • Paul –
        as I said elsewhere on this thread “no AGW in response to doubling of CO2″ is within the subjective probability estimates of climate scientists.

        Which climate scientists, with the exception of those few who are “known skeptics”, would publicly admit the possibility? I think you overstate the case.

  35. Tomas Milanovic

    Judith

    I am unconfortable with this terminology of “black swans” or “dragon kings” because it introduces something that can’t be dealt with by the science.
    Namely a notion of unimagined or unrecognised.
    That’s why there may be an impression of confusion in the discussion above.
    It is because you cannot know (now) what is unknown (now).
    By definition.

    I will take an example of the time when in another life I have been working on safety systems in nuclear power plants.
    One of the requirements for design is to minimize damages in case of an impact by a large plane.
    The plane plays only a role because it provides an envelope for energy , mass and momentum that are needed to do concretely the design.

    Now when we do that we have not a single clue about the probability of such an event and actually nobody cares.
    It can be estimated by [surface of all nuclear plants/surface of the Earth]x[number of plane crashes in a year / number of planes flying in a year].
    This gives ~10^-12 .
    Despite the fact that this number may be off by several orders of magnitude, it is a ludicrously small number.
    It means that IN AVERAGE such an event would happen once every 1000 billions years what is about 100 times the life duration of the solar system!
    Or it could happen tomorrow …

    But such an event is certainly not a black swan. It is perfectly described and anticipated. We only don’t know what is its probability and when it would happen.

    Now an event where an ET from a civilisation with largely superior technology would teleport to the control room and stop the main pump while letting the reactor in full power would be a black swan because teleportation has never been considered for safety design.
    Or at least it was the case untill the previous sentence has been written.
    Now it became just another event with a ludicrously low probability which, by the way, would not be qualitatively different from the 10^-12 considered above.
    When the time unit becomes the age of the Universe, all notions of probability loose any practical meaning.

    I made this comment only to make a point that in practical considerations (e.g those considerations that lead to investments and actions) concerning ludicrously meaningless low probabilities, the only limit is the imagination.
    And the difficulty that all safety system designers well know is that there is no limit to imagination.
    The human being is PACKED full of black swans – just think about the last case when somebody blew himself up and killed 50 people or found another creative and unprecedented way to do damage.

    There is an infinity of imaginable events with probabilities of order of 10^-10.
    So you need somebody who comes and says “OK guys , now it’s enough. We’ll stick just with the plane and the unprecedented earthquake.”

    In other words the final decision about the scenarios (e.g some domain of parameter values for the design) is not and cannot be science.
    The scientist, experts and engineers left to themselves would find new unprecedented events all with same ludicrously low probabilities forever.
    So there always is an arbitrary decision mostly triggered by the boredom of the one in charge who can’t afford to listen to new scenarios forever.

    • Hi Tomas,

      Dragon Kings are defined as the extreme variation associated with bifurcation. As in – Climate tipping as a noisy bifurcation: a predictive technique http://www.ucl.ac.uk/~ucess21/%5B201%5D%20IMA%202011%20(Ys).pdf

      Nothing to do with black swans at all. It has to do with slowing down and noisy bifurcation. But Dragon Kings is a better descriptive – it gives an impression of the power and significance of these events in climate.

      Black swans are defined in the Popperian sense as evidence that a theory may be falsified by new data – e.g. the existence of black swans. Taleb uses it with an additional condition that black swan events are unpredictable and change society in major ways. I think this is misapplied here to events like storms and cyclones – which cannot be truly unpredictable. I quantify 1 in 10,000 year storms all the time using a log-Pearson type 3 statistical distribution.

      Cheers

  36. Jonathan Gilligan

    Judith: What is your sense of the difference between Taleb’s Black Swans and Martin Weitzman’s fat tail climate risk? My reading of Taleb is he’s saying one of the big problems with black swans is that people assume a normal distribution and thus underestimate the probability of extreme events.

    Consider too the connection between this line of thought and Max Bazerman and Michael Watkins’s notion in “Predictable Surprises” that anthropogenic climate change is a predictable surprise in the same way that the Enron collapse and the 9/11 attack on the WTC were predictable surprises.

  37. “Black swans,” by definition cannot be anticipated. From the perspective of an intellectual construct, their probability is essentially random. If they are imagined, they will be conceptualized only by individuals that society will view as “crackpots.” Because of their rarity and systemic impact, no resaonable probability can be calculated. Even if conceptualized, their very rarity will exhaust the patience of those who rely upon probabilistic models.

    • Diogenes,

      I’d argue that Black Swans can be anticipated, though not predicted. Some of the examples Taleb gives (World War I, 9/11) were certainly imaginable beforehand. I agree that the random nature prevents calculation of a probability, but that is a weakness of over-reliance on probabilities rather than Taleb’s theory.

      • WW1 is not really a Black Swan, but not because it was unimaginable. The fact that a phenomenon is imaginable is surely not the point – those who thought all swans white were not incapable of imagining swans of other colours, they just believed they had seen all the swans there were to see, and since that they were all white, concluded that all swans were white. Come to think of it, I can imagine a green swan – so what?

        In the case of WW1, where white swans represent the persistence of peace, and black ones an outbreak of global conflict, black swans had existed before in the form of the Napoleonic Wars, so the question of imagining them doesn’t really arise. War was indeed imagined before August 1914, but as Niall Ferguson points out in “War of the World”, those best-placed and with the most to gain from predicting it, the financiers, continued to think it unthinkable until a few days before it broke. He cites correspondence Rothschild family which amply bears this out, and points to the absence in the bond market – a reliable canary for conflict – of any expectation of war almost until its eve.

      • As you noted, there had been global conflicts previously (from the mid 1700’s on through Waterloo), but the speed and almost automated manner in which the beligerants went to war in 1914 was unprecedented, though hardly unimaginable. While the financial world was caught flat footed, the Central Powers fully expected the invasion of Serbia to trigger either a general war or a break up of the Triple Entente. What was a black swan for the various governments was the evolution of the military situation. While there had been hints of how industrialization would affect warfare, a completely static war was unprecedented.

        Come to think of it, I can imagine a green swan – so what?

        This, in my opinion, is the crux. An over-reliance on what we think we know leads us to discount the unprecedented. We fail to acknowledge that precedents were once unprecedented.

      • In a way WW1 was more a transition phase, thus a dragon king.
        the world was coming from a feodal world, with small independant powers.
        but the powers were getting bigger, and then start to couple together, in response to coupling of usual enemy.
        it was a phase transition, self catalyzed.
        and the was was the consequence of that coupling with the sparks of an unavoidable trouble between long time enemies.

        the typical dragonking.

        black swan are more like a chain of problems that propagate more often than expected, but not always (unlike dragonkings).

        in climate a blackswan is something uncommon that we already know as rare event… same with earthquakes… we have statistics, but they are not gaussian.

        an earthquake M9 is a blackswan.

        but 20 small earthquakes M6 correlated all over the planet, it is a dragonking…

        however for some here the definition of black swan as really unexpected , might be : a M10 earthquake… in theory impossible, like was the M9 in front of Sendai according to previous theories of that zone.

        a agree that my classification is different.
        for me blackswan and dragonking are statistics event in a dynamic system… they are all possible in theory, but often modelized wrongly as improbable. when you see some, you can modify your model, so you capture their true probability.

        impossible events that happens, are … reason to change your theory, and recompute all.

  38. • This has been a great Thread, although it begins by giving too much prominence to Taleb’s version of Popper’s original Black Swan finding, where although citing Popper in general quite often, he somehow manages to omit Popper’s actual Black Swan statement, which is really the same as Einstein’s No amount of experimentation can ever prove me right; a single experiment can prove me wrong.
    • Setting Taleb’s version to one side, let’s get back to Popper&Einstein, and apply them to the IPCC’s central hypothesis, that “most” of the temperature change observed over the last century is attributable to the build-up in the atmosphere of anthropogenic emissions of CO2e, (Hegerl Zwiers et al in Solomon et al. 2007) of which CO2 is by far the largest in both volume terms and rate of growth.
    • In a currently under peer-review paper (known to JAC) I have shown that the hypothesis fails the Einstein-Popper test at three selected locations, Barrow in Alaska, New York, and Hilo/Mauna Loa near the Equator in Hawaii, at all three of which it can easily be shown, using only NOAA-ESRL and CDIAC data, that changes in the atmospheric concentration of carbon dioxide (hereafter denoted [CO2]) have ZERO statistically significant impact on changes in temperature (whether Tmax, Tmin. or Tmean). That is THREE Black Swans.
    • But if they are not enough, here are my results for JAC’s own Georgia Tech at Atlanta:
    1. Trends in changes in temperature at Atlanta 1960-2006 (when the NOAA-ESRL series mysteriously terminates):
    Tmax: y = -0.0011x + 0.0186
    R² = 0.0004
    Tmin: y = 0.001x + 0.0043
    R² = 0.0004
    Tmean: y = -0.0002x + 0.0118
    R² = 2E-05
    2. It would seem that 47 years of data at Atlanta provide no support for ANY rising trend in temperature despite the IPCC’s claims on the basis of the monotonic increases in CO2.
    3. But to be fair, let’s consider in more detail. First, Tmax. Regression of changes in Atlanta’s Tmax from 1960 to 2006 against Changes in solar SURFACE radiation (AVGLOBAL in the ESRL data), precipitable water (“H2O”), and atmospheric CO2 (“[CO2]”, per CDIAC at Mauna Loa) shows an R2 of 0.40, and t-statistics respectively of 5.62, 3.49, and 0.018. As not a single IPCC climate scientist cited as lead authors etc of the IPCC’s Solomon et al. 2007 knows what a t-stat is, let me explain that it is the respective coefficient divided by its standard error, so if t = <2, it is rubbish, as is the case for changes in [CO2] here, whereas the t-stats for changes in [H2O] and in situ solar radiation (“AVGLO”) are hugely significant. Another pair of black swans!
    4. Next, Tmin. Strangely, as Solomon et al. clearly are not aware that the sun does not shine at night, whereas the opacity (OPQ, a term unknown to the IPCC) of the sky becomes relevant, if we replace AVGLO by OPQ, then we have these results, that OPQ has a larger role than [CO2], but without being statistically significant, whereas the main player as before is the ESRL’s “precipitable water”, i.e., atmospheric water vapor, denoted here as [H2O], hugely statistically significant (t stat = 3.39, well above the benchmark 2.0). Another black swan, showing that changes in [CO2] are irrelevant to any explanation of minium temperature changes in Atlanta between 1960 and 2006.
    5. Now how do I explain these results? The IPCC’s Solomon et al. 2007 totally reject any role for natural changes in [H2O] as UPFRONT contributors to radiative forcing – for them, including their Forster & Ramaswamy et al. and Hegerl & Zwiers et al. the only role for atmospheric water vapor is as a positive feedback from rising temperature allegedly caused by rising [CO2] p- but this is absurd. The sun’s radiative forcing is 342 W/sq.meter (Kiehl & Trenberth) whereas the TOTAL RF from GHGs in Forster & Ramaswamy in Solomon et al 2007:141) was only 2.66 in 2005. Tails wag the dogs? Really? Not mine!
    6. Alternatively, Solomon et al. claim that the Global Temperature Increase of 0.7oC between 1900 and 2000 (i.e. 0.007oC p.a.) was enough to increase evaporation and thereby [H2O] by more than any annual variation in solar SURFACE radiation. Really? A rise in temperature of 0.007oC in say 1995 was enough to increase evaporation and thereby [H2O] by more than either natural or DIRECT human variations in solar radiation?
    7. In addition the Solomon et al lead authors (Forster & Ramaswamy, Hegerl and Zwiers, and some hundreds of their acolytes, although mostly they cite only themselves, ad nauseam) totally ignore that the process of combustion of hydrocarbons (terms not in the Index to Solomon et al) produces BOTH H2O as well as CO2 AND Energy, used to produce STEAM to generate the power that produces our electricity. What is Steam? The IPCC’s Susan Solomon has never heard of it (it is not in the Index to Solomon et al.). But the annual production of Steam by the worlds’ power industries is much more than TEN times their output of CO2, and has DOUBLE the latter’s warming potential (as well as raising the world’s annual rainfall).
    8. Truly, the defining characteristic of the IPCC’s Climate Change 2007. The Physical Science Basis (sic) (Solomon et al. 2007) is the complete absence of any science – it never shows the chemical formula for combustion of a hydrocarbon fuel, yet that combustion is what it demonises, even though it has been the basis for the huge improvement in global living standards since 1750.
    9. Susan Solomon claims to have trained as a chemist. I find that hard to believe. Be that as it may, some of us are preparing a class action against her and her fellow Nobel winners for a gigantic Ponzi that far exceeds Bernard Madoff’s in its wealth destruction potential.

    • Seven point two billion or so, and rising in that class. Include in the class all future generations, please.
      ==========

  39. If anything has become clear in this thread, it is that the concept of black swan is problematic as everybody has his own meaning for it and these meanings are widely different. The Dragon King must be even worse, as it’s unknown to almost everybody. Taleb’s book has been a bestseller while only few know about Sornette.

    Metaphors may have a communicative value, if they are understood in the same way by most, but it’s clear by now that neither of these concepts passes that test, although most seem to think that they know the right meaning of “Black Swan”.

    That doesn’t imply that the rare events or individually highly improbable outcomes would not sum to something very important, but obviously that must be explained more directly. Taleb’s book may be a worthy reference as may be Sornette’s article, but saying “Black Swan” of “Dragon King” does not purvey the message, but leads rather to misunderstanding or semantic argumentation as in this thread.

    • Pekka Pirilä

      It may be true that people have different ideas of exactly what a “black swan” really is, as you point out..

      But here is a good example.

      Some time after the mid-19th century there was concern about the rapidly increasing number of horse carriages in London (the same was apparently also the case in New York City). The worry was that, if the rate of new carriages continued or even accelerated, London would be covered in two meters of horse manure by the year 1930.

      The “black swan” here was the automobile.

      Max

    • Pekka,

      Dragon Kings are defined quite closely as being the variability associated with chaotic bifurcation. As I have said here – https://judithcurry.com/2011/05/02/anticipating-the-climate-black-swan/#comment-66503

      It has a real physical meaning and the confusion comes from not understanding the underlying physics rather than from looseness of definition. A correct understanding of Dragon Kings is essentially for understanding the most critical underlying processes in the theoretical physics of both models and climate – that of dynamical complexity. This is observed as abrupt climate change – which is defined as rapid and non-linear departures from a state.

      The concept is defined in Sornette – but I have linked to another paper in my response to Tomas where it is called noisy bifurcation.

      Cheers

      • Rob,
        Didier Sornette has given a rather precise definition for what he calls a dragon-king, but nobody has the power to control, how people will use a metaphor, if and when it gains popularity. The problem with metaphors is that people will not go to learn the definition, but form their own picture based on the impression the metaphor gives.

      • The paper of Sornette contains also a large number of pictures that demonstrate, what I have written in other messages: The power laws are never precise, they extend never over the whole range, but are always only parameterizations approximately valid over a limited range.

        The power laws are a demonstration that normal distribution is not generally valid. They tell, how common are outliers in comparison with predictions of the normal distribution. It’s paradoxical that Sornette and others take power laws so seriously that they start do discuss outliers with respect to power laws. How can one have outliers compared to a law that is not a law at all, but only a empirical parameterization.

      • There are 2 separate problems. The problem of the misunderstanding the terminology which is a function of not understanding the theory. It is the wider problem people understanding the technical details and understanding them in terms of dictionary definitions.

        The other is the power laws. There are characteristic shapes to empirical distributions of rainfall for instance – described most commonly by a log-Pearson type 3 distribution to obtain values for extreme events for which, of course, we lack data. What Sornette talks about is where there are defined power distributions of this sort is results that are conspicuously different from what the distribution would lead you to expect.

        You go to the core of the paper – in respect to Paris at least – and suggest that the methodology is itself paradoxical.

      • One has to wonder what these scientists smoked in their student days.

  40. Correction: At #6 I said “A rise in temperature of 0.007oC in say 1995 was enough to increase evaporation and thereby [H2O] by more than either natural or DIRECT human variations in solar radiation?”

    Please replace”human variations in solar radiation” by “increases in human production of [H2O]”.

    Apologies! – it’s late (midnight) here in Canberra, Australia.

  41. “The important point you should take away from Taleb is that experts are not nearly as smart as they think they are. See, e.g. Teltlock.”

    This is exactly right. The recent cold, snowy winters are black swans if you listened to the “experts,” like the featherheads at UKMET with their bogus warm-biased models, but if you’ve the brains and judgment to find the few guys who really know what they’re talking about, Bastardi. D’Aleo etc., then you weren’t surprised.

    I’ve been making the connection for a long time now, between the “experts” who failed to see the housing bubble (how blind was that!) and their AGW counterparts who are equally clueless about the coming cold…

    Warming world? I think not.

  42. Beth Cooper

    Black swans are accommodated in CAGW by what Popper refers to as theory innoculation. ‘Global warming’ is simply renamed as ‘climate disruption.’

  43. Ken Lydell

    Much ado about nothing in this as far as I can see. The way in which one classifies an event in no way changes the nature of the event. Events that are both extreme and rare can be characterized as extreme and rare. Unanticipated events are just that: unanticipated by some identifiable group of people. The term “unimaginable” is essentially meaningless in the absence of an objective measure of imaginability. In short, Black Swans and Dragon Kings are at best amusing metaphors. Attaching one of these labels to some climate event with extreme properties explains nothing, helps predict nothing and has no practical use.

  44. I’m finding the distinction between “black swans” and “the black swan theory” confusing, and perhaps inherently problematic.

    The original point of the black swan concept was really just a restatement of David Hume’s observation, that science doesn’t really make any sense, because no matter how many times we observe something, it doesn’t mean it’s going to happen that way again the next time. But even David Hume admitted that, however little sense it makes, science seems to work. The nature of this reconciliation is sometimes accepted as a statistical proposition, i.e., predictions based on induction will sometimes be wrong, but they’re right often enough to bet on them–a bit like choosing to hit on 15. But it’s really rooted in a disagreement about an axiom. Science operates from the axiom that given identical initial conditions, you’ll always get the same result. (Quantum uncertainty was a problem, but at least some scientists believe it’s been reconciled.) But that’s an axiom–it is unproved, and, Hume notes, can never be proved.

    But Dr. Curry specifically distinguished “black swans” from the “black swan theory.” As I understand it, they are defined by their effect on the collective thinking of a community, rather than by the fact that they falsify some previously accepted inductive conclusion. So my question is, defined this way, what is the utility of the concept? As one poster noted, with this subjective definition, it seems like the test for whether something is a black swan is run on the audience in question, rather than the information itself.

    Dr. Curry also calls them “outliers,” which seems to me to begin to confuse them with dragon kings (or, at least, my understanding of them). On the other hand, it seems like it’s quite possible to be both. If the weather was a significant deviation from the weather we expected, but within the range of credible statistical noise, it was a dragon king. If one believed such weather was impossible given AGW, it was a black swan. (?)

  45. 1. What is fragile should break early while it is still small. Nothing should ever become too big to fail.

    The above is about as wrong headed as it gets.

  46. “The above is about as wrong headed as it gets.”

    I just don’t see the point in dropping little assertion bombs without some sort of rational defense of same.

  47. “The unexpected cold, snowy northern hemisphere winter (2010/2011) and the flooding (2011) in drought stricken Queensland highlights how our expectations of extreme events in a warmer world can be soundly trumped by natural variability of weather processes.”

    To me, a climate Black Swan would be something like the nonlinear collapse of the Greenland ice sheet.

    • Funny you should mention the increased melting of Greenland as a black swan since I was just trying to come up with a couple and that was one I was thinking of that would be a black swan for me. The one I was thinking of that might be a black swan for you is the AMO goes negative and the Arctic sea ice recovers to previous levels.

      • I would not consider that a Black Swan – unless the ice recovered far in excess of what would normally be expected under cooling conditions.

      • It would be rather difficult to see increasing sea ice from the AMO and achieve an ice free Arctic ocean by 2050. Would you say the idea of an ice free Arctic ocean by 2050 is a black swan or would you just say I am not going far enough out on a limb to have these qualify as black swans?

  48. “black swan”: noun, deep philosophical insight developed by the most sophisticated academic analysis to say that “sh*t happens.”

  49. ferd berple

    Black Swans exist because human beings largely assume that events in the unverse are random, that over time, “things will average out”. So, if you toss a coin, you may get a run of heads or a run of tails, but over time things will average out to 50-50.

    Does this hold for climate? Well it could be argued that some years are hot and some are cold, so over time they should average out. Thus, climate should be easier to predict than weather. However, it turns out that this is statistical nonsense.

    When you look at temperatue, it is not like a coin toss. A coin has only certain states, heads and tails. This makes it predictable. We assume climate is predictable because we have made a mistake. We assumed that temperature has only two states, warmer and colder, and thus will average out.

    Temperature however can go up and down, and the longer the period you sample, the bigger temperature change you are likely to see. This gives rise to a fractal distribution, which has a very interesting property.

    Fractal disributions are scale independent. You cannot tell if it is for 10 years of 10 million years, it will display a similar appearance. This is extremely important because it means that that climate is equally predictable at all time scales. It does not become more preditable the as the length of time increases.

    This is why Black Swans exist. We assume that climate (nature) becomes more predictable over time. The longer we go without seeing a black swan, the more we assume that they don’t exist.

    Nature doesn’t work that way. Black swans did not always exists. Even if a black swan doesn’t exist today, the longer we wait, the more likely Nature is to create one. Expect the unexpected.

    • fred berple

      Interesting comment.

      Taleb points out that the longer-range a prediction is, the greater the chance for error.

      The “heads/tails” analogy to try to justify an opposite conclusion (i.e. “climate in 100 years is easier to predict than tomorrow’s weather”) is obviously flawed, for the reasons you point out.

      But I have also heard another mathematical explanation, namely: climate is not really the integral, or average, of weather – it is the sum.

      Not being a “climate insider” (or statistician), I have not pondered too much on this.

      Maybe an “insider” would like to comment.

      Max

      • John Kannarr

        “But I have also heard another mathematical explanation, namely: climate is not really the integral, or average, of weather – it is the sum.”

        I am definitely not a climate insider. I can’t even begin to imagine what that statement might mean!

      • ferd berple

        Imagine that you were to bet if a fair coin would turn up heads or tails, and the amount you bet was randomly chosen between 0 and the house limit, and the house limit slowly increased the longer you played. climate (average temperature) is your accumulated winnings.

        Now predict how much you would be up or down at any point in time. You can’t. Even with a fair coin there is no way to predict your actual winnings.

    • Another way of saying that – and this is really the climate/weather distinction problem – is that low frequencies are signal and high frequencies are noise. From this, they conclude that there’s a corner frequency somewhere, and everything above this corner frequency is weather noise, and everything below this corner frequency is climate signal.

      Of course if it’s really fractal, this is false. Conservation of energy argues that at some point this has to be true, but the planet isn’t an RC network.

    • But there are no real fractals in nature. They exist only as mathematical artifacts. Many phenomena in nature show similarities at several different scales or over a range of scales, but never at all scales and never with absolute precision.

      The fact that some phenomena show fractal like properties over a range of scales is never a proof that some other phenomena behave the same way.

      It is not excluded that climate is naturally highly variable over wide range of time scales, but it’s not proved either.

      • Pekka, this is true of most, if not all, math. For example, nothing in nature is mathematically continuous but calculus still works in many cases. I think fractal (or “strange”) statistics play a major role in climate.

      • Noting that the normal distribution fails miserably to describe many probabilities and that the tails are often much fatter has been important in many fields and applies certainly also to many issues related to climate. This is the most solid and important observation of Mandelbrot, Taleb and many others. The normal distribution has been used so widely, because it is so easy to apply and also because there remain very many applications, where it presents the probabilities well.

        The conditions that lead to normal distribution are well known and easy to describe, but conditions that would lead with high precision to power laws are diverse and complex. The power law is, however, flexible enough to fit tails of many PDF’s with satisfactory accuracy. This is an empirical observation with limited fundamental importance. There are certain specific phenomena, where power laws have fundamental basis, e.g. critical phenomena on physics, but mostly they are just useful parameterizations of empirical data or results of more complex models.

        Such parameterizations are accurate enough over limited ranges and fail often badly outside those ranges. In most cases they are used only for the tail without any attempt to describe the whole distribution with a single formula. Even the tails have usually a cutoff at some point where the power law stops working, but that may be at a very low level of probability.

        The power laws are useful as an approximate description of reality, but they are usually not the reason that can be used as evidence for some arguments.

  50. Nebuchadnezzar

    I don’t want to sound horribly cynical here – but, after so many comments I’m not sure that is really possible – but wouldn’t it be easier and better value for your readers if you were to say what *you* think rather than reheating other people’s metaphors?

  51. Lord Frijoles

    Having read all the comments carefully, I’d like to make an attempt to clarify the “black swan” theory and the events that allegedly characterise it through the lens of Popperian epistemology.
    It is my understanding that Popper advances the idea that any scientific theory (i.e.; a testable theory) should deny the existence of particular and specific events/phenomena. These specific events/phenomena in turn become the “empirical content” of the proposed theory, because if evidence of the existence of these events/phenomena was found, it would, in essence, falsify the proposed theory that negates the existence of such specific events/phenomena. Popper exemplifies this with the law of conservation of energy that states that no perpetual motion machine can exist. Thus, if upon verification and scrutiny a statement such as “in XYZ part of the world there is a perpetual motion machine” was found to be accurate, it would falsify the law of conservation of energy.
    Now, with respect to the proponents of the black swan theory, I don’t understand the examples they give to support their theory. When they talk about black swans, are they talking about potential “falsifiers” of commonly-accepted theories/laws/knowledge? Or are they talking about rare, huge and high-impact events that, though not very common, have previously taken place but were not taken into account by modellers/decision-makers when they made their predictions? If it is the former, then the theory of “black swans” doesn’t offer anything new in my opinion. If it is the latter, however, the mere use of “black swans” to label their theory is confusing, to say the least, because the “black swan” example is a well-known metaphor used by Popper and other philosophers of science when explaining the occurrence of events that go against the predictions of well-established theories/laws.
    Just my two cents.

  52. Lord Frijoles

    Having read all the comments carefully, I’d like to make an attempt to clarify the “black swan” theory and the events that allegedly characterise it through the lens of Popperian epistemology.

    It is my understanding that Popper advances the idea that any scientific theory (i.e.; a testable theory) should deny the existence of particular and specific events/phenomena. These specific events/phenomena in turn become the “empirical content” of the proposed theory, because if evidence of the existence of these events/phenomena was found, it would, in essence, falsify the proposed theory that negates the existence of such specific events/phenomena.

    Popper exemplifies his theory with the law of conservation of energy that states that no perpetual motion machine can exist. Thus, if upon verification and scrutiny a statement such as “in XYZ part of the world there is a perpetual motion machine” was found to be accurate, it would falsify the law of conservation of energy.

    Now, with respect to the proponents of the black swan theory, I don’t understand the examples they give to support their theory. When they talk about black swans, are they talking about potential “falsifiers” of commonly-accepted theories/laws/knowledge? Or are they talking about rare, huge and high-impact events that, though not very common, have previously taken place but were not taken into account by modellers/decision-makers when they made their predictions? If it is the former, then the theory of “black swans” doesn’t offer anything new in my opinion. If it is the latter, however, the mere use of “black swans” to label their theory is confusing, to say the least, because the “black swan” example is a well-known metaphor used by Popper and other philosophers of science when explaining the occurrence of events that go against the predictions of well-established theories/laws.

    Just my two cents.

    • Taleb made the concept of black swans widely known, but his black swans have very little to do with Popper’s black swans. Taleb starts with the invalidation of the hypothesis “all swans are white”, but the rest of the book is about the importance of rare events, not philosophy of science.

      The opening post of this thread referred to Taleb, not Popper. All comments related Popper’s black swans in this thread are totally unrelated to the opening posting.

      • Lord Frijoles

        Thanks for the comment.

        However, in my opinion, much of the confusion (just have a look at the posts on this thread) regarding so-called “black swans” events are due to the extremely poor foundations of so-called “black swan” theory. More specifically, the vagueness of the statements that support such a theory and the “black swans” events that would characterise it just add more confusion to the debate and are of little help to understand why climate science and the policies derived from it are in the mess they are right now.

        For example, let us suppose that perpetual motion machines can actually be built. If an engineer today designed a machine based on the law of conservation of energy (“no perpetual motion machines can be built”), would you blame the engineer if in the future the machine failed because you can actually build perpetual motion machines? I can’t see how you’d blame him, because there is no evidence right now that such machines can be built. Thus, it would be unreasonable to think the engineer should have included such an event in his model. A timely question here would be: Would a perpetual motion machine be considered a “black swan” event?

        What I have read is that recent natural and financial disasters could/should have been foreseen because (though rarely) those events have happened in the past and even with certain regularity. So it is entirely reasonable to criticise modellers/decision-makers for not including those (rare) events in their models and predictions. The question again is: should these occurrences be labelled as “black swans”?

        Now, consider the two cases I just laid out: a perpetual motion machine, and recent financial and natural disasters. In both cases, the policies/models that were made/built excluding those events turned out to be wrong. Evidently, in the case of the engineer, it’d be unreasonable to expect any designer to build models that would include the possibility of perpetual motion machines, as there is no evidence of them. But in the case of financial and natural disasters, we should expect modellers and decision-makers to include them as there is evidence of their occurrence.

        So, going back to so-called “black swan” theory and the events that characterise it, how does it help us to understand occurrences such as the two I described above? I’d really like to know, because from my perspective, I can’t see how this theory (if it is indeed a theory) can be of any help at this moment.

        Thanks in advance!

      • In my view there is no Black Swan theory, not in the spirit of Popper nor in the spirit of Taleb. Popper used black swans as an example of refutation, while Talebs’s book doesn’t even try to present any specific theory. He discusses rather weaknesses of too simplistic ways of looking at probabilities, and presents examples and narratives of, how the too narrow ways of looking ant the world may lead us astray.

        Taleb’s discussion has much in common with Benois Mandelbrot, who has discussed also more precisely defined mathematical approaches, but even Mandelbrot is often forced to move from theory to narratives and vague descriptions of his ideas.

        Concerning our capabilities of making predictions both Taleb and Mandelbrot tell much more about the impossibility of making predictions of many types than about, what can be predicted in spite of the limitations.

      • I think Taleb would have done better to talk about a platypus than a black swan. A black swan is no less evolutionarily plausible than a white swan. A platypus, OTOH was initially considered so implausible, it was assumed to be a hoax.

      • A black swan is no less evolutionarily plausible than a white swan.

        Which is, I believe, Taleb’s point.

      • I understood it to mean that platypus happens. Less plausible things happen all the time. Which just goes to show that no two people on this thread understand the same thing. This whole subject is a room full of concave mirrors.

    • I think that makes four cents. But still good value.

  53. SteveGinIL

    Much more than the US or European cold spells and winter storms – all of which were within or only slightly outside previous record weather – was the South American weather. SA had freezing temps well into the tropics.

    In addition, the weather in Australia the day before their summer solstice was more mind-boggling, as no place in the country had an average temperature as high as 40F and there was snow as close to the Equator as Havana – the day before summer started.

  54. Judith, all this discussion about what is and is not a Black Swan or Dragon King is missing the mark about your original post.

    The Black Swan Theory was coined because Black- Scholes Option models, upon which trillions of dollars in derivatives were based, started failing their owners when the financial universe changed. What were insignificant forces whose effects could be neglected in the B-S calculations quickly became very significant factors in the economy which swamped B-S assumptions. What the models said would be very improbably started to happen daily.

    In reference to your paper and Climate Models:
    If the climate models work, they work.
    Can the climate models fail us?
    Can the world turn out differently than the climate models predict? If so, what can render the decisions based up climate models useless and/or dangerous?

    At this point, you look at the assumptions of the Climate Models and list those factors not included or outside the commonly accepted ranges.

    For instance:
    The Sun is really a variable star. Its output varies by +/- 2% over the past 20,000 years.
    Cloud Feedback is negative.
    Global Average Temperature is a useless proxy for climate change. (Example: Minimum Winter Averages are far more important.)
    Global Warming over the past 30 years was primarily attributed to cleaning up particulate emissions, not CO2 increases.

    These are potential IPCC Climate Model Black Swans: Things that the MODELS assume to be impossible or unimportant.

    When you get back to the question of how does this affect policy, you must really expand your decision tree to combine where action meets Black Swan event. This is fruit for another post, but in short the key question is
    What if we do nothing?
    What if we do something on preventing the wrong thing?
    What if mitigation tomorrow is much cheaper than prevention today?

    Black Swans must be linked to the MODELS that FAIL to describe them. The Earth is what it is. Black Swans are only important in how our simplified, tractable, experimental models of the Earth may be dangerously in error.

  55. SteveGinIL

    This Black Swan concept is just another term for “paradigm busters.” I have noticed these for a long time, in several fields. At least that is how I see it when scientists from astronomers to paleontologists are quoted in the media as saying something like, “Wow. We never expected anything like this was possible – now we are going to have to go back to the drawing board, because this means everything we thought was true about this area of study needs to be rethought.”

    I refer to these moments as “Science Does It Again,” and have seen them for decades, but have only recently thought to begin a folder in my browser for such paradigm busters.

    It is not necessary that such paradigm busters be something along the lines of falsifying the Law of Conservation of Energy. Though that should be considered one, that is completely setting the bar too high.

    Black Swans in recent times would be the Hobbits and the discovery in Russia about a year ago of a FOURTH homo sapiens.

    The biggest in our time may have been Shoemaker-Levy-9, the disintegrated comet that impacted Jupiter several times in 1994. That put the lie to the uniformitarian concept that all the big impacts in the Solar System happened eons ago and that such impacts in the time of Man are once-in-million-year events. When the plume in Jupiter’s 8G gravity was as large as the entire Earth, geologists and astronomers knew our safe little world wasn’t safe at all, and that catastrophe could rain from the sky at any moment.

    The second biggest was Kerry Mullis’ discovery of PCR, the cell-multiplying miracle that was the biological equivalent of a nuclear chain reaction. From it came DNA testing, the human genome project, and the genetic mapping of the history of mankind, to name a few.

    The third – and probably the biggest of all time – was stem cells. What miracles will come from this we hardly know yet, but it sure wasn’t on the map 15 years ago. I have a personal friend who had no cartilage in one knee and ACL and MCL damage in both knees. This woman got stem cell treatment in the last 6 months, on both knees as well as her carpal tunnel syndrome. Within DAYS, her knees were functional at about 85%, and are now as good as when she was in her 20s.

    The fourth Black Swan/paradigm buster was the Monte Verde site, which shot the Clovis Barrier to smithereens. This opened up the study of man in the Americas to pre-Clovis, something that had not been allowed into the peer-review literature. It now turns out that there were (so far) FIVE human incursions into the new world, based on genetic studies done (made possible thanks to Mullis’ PCR).

    It is my take on things that we have Black Swans every year in science, or very nearly so, and some years more than one.

    Black Swans are just clay pigeons, because science makes hard and fast declarations about things, and instead of understanding that these declarations are only tentative, they act as if they are gospel. The AGW warming concept was declared back in the 1980s, with hardly any solid evidence – with too little to make such a claim. We have been stuck with it ever since, and only thanks to the “skeptics” and Climategate has the world begun to recognize that the claim was WAY premature. It does not qualify as a Black Swan, since it claimed ascendancy only by ignoring conflicting evidence and blackballing climate scientists who had such evidence. That does not qualify as a Black Swan – only as politics. But it IS fun to jab at them in their fallacious state of mind. They set themselves up.

    Setting ones self up for pot shots does not make a Black Swan. Anyone with a tinfoil hat can do that.

  56. I propose a new thread – How Many Black Swans Can Dance on the Head of a Pin?

    But seriously folks…. This thread is a good example of why policy should be informed by academics, but not formed by them. The concept of a black swan may be a very useful metaphor to help discuss the issues involved in understanding the limits of human knowledge, particularly when dealing with complex, chaotic systems like the climate. But when we get diverted by whose black swan is a better metaphor than someone else’s, we may have reached the point of diminishing returns.

    One of Dr. Curry’s constant refrains is that the climate science consensus scientists have overstated the degree of their certainty. To the extent that policy makers (such as those controlling the dam in Queensland) are convinced of that certainty, foreseeable severe weather events become unforeseeable (to those policy makers), with sometimes catastrophic consequences.

    Events like massive asteroid strikes and extinction level volcanic eruptions don’t seem like something worth spending a whole lot of time worrying about. But the failure to properly use a dam for flood mitigation, or the failure to properly acquire sufficient available energy resources to cope with a potentially severe winter, are real world consequences of the self imposed limitations on planning that result from too great a level of certainty.

    Whether you call them black swans or blue ducks, the larger point seems to be that a certain amount of humility regarding what we think we know, may help us prepare for the things we do not (or cannot) know. Simply being aware of the fact that there is much that can happen that we cannot predict, could help prevent some similar future policy errors.

    Perhaps
    “Such variability has important implications for the assessment of dangerous climate change and for reducing vulnerability to weather disasters.”
    could better be stated
    “Such variability has important implications for the assertions of levels of certainty unsupported by the current level of scientific knowledge regarding climate change. Correctly accounting for uncertainty, and accepting the possibility of extreme weather events that may be counterintuitive to current climate predictions, could reduce future vulnerability to unexpected weather disasters.”

    Or not.

    • There’s a reason why that is. In academia, you get points for novelty, sophistication, counterintuitivity, creativity and other such things. In the rest of the economy, the premium is on being right, and everything else takes a back seat to that. Being right is pretty far down the list in the academic world, partially because even in the hard sciences, you’ll probably be dead before the question is actually settled. So they use all these other things as proxies for value.

      When people elsewhere in the economy can collect their bonuses and move on before the truth is revealed, we get such things as the real estate bubble. Nothing ever works well when people can’t be judged by the absolute consequences of their conclusions and decisions. Those are the last people who should be determining policy.

  57. After having read the many comments here, I see that the discussion is “getting wrapped around the axle” on the definition of a “black swan”, rather than staying on the main topic of how uncertainties and potential major outliers in our knowledge of climate science should affect our conclusions of what is likely to be the human impact on climate and what should be done about it.

    As I understood it [correct me if I’m wrong, Judith] Judith has suggested that “fully exploring the uncertainty and the possibility for black swans” (since they are impossible to truly identify or quantify by definition) might be a good start.

    The framework associated with setting a CO2 stabilization target focuses research and analysis on using expert judgment to identify a most likely value of sensitivity/ warming and narrowing the range of expected values, rather than fully exploring the uncertainty and the possibility for black swans (Taleb 2007) and dragon kings (Sornette 2009). The concept of imaginable surprise was discussed in the Moss-Schneider uncertainty guidance documentation, but consideration of such possibilities seems largely to have been ignored by the AR4 report. A key issue is to identify potential black swans in natural climate variation under no human influence, over time scales of one to two centuries.

    Some of the posters here have stumbled over the “black swan” reference, but to me this means simply acknowledging [again, correct me if I’m wrong here, Judith] that we (climate science) do not have enough information today to give any meaningful advice to policymakers regarding the need to set CO2 stabilization targets.

    This is because our level of knowledge today is so rudimentary that we cannot make any reasonable estimates of what the impact on climate of human CO2 emissions has been in the past or will be in the future.

    As a result, we should best advise policymakers as well as the public that now is not the time for urgent action.

    It is, instead, the time to attempt to get more clarity on the many uncertainties and unknowns we currently face, in particular regarding “natural climate variation under no human influence, over time scales of one to two centuries”, an unanswered question, which has been largely ignored in the AR4 report.

    I think Judith has expressed this very well in her testimony to the Baird committee of U.S. Congress last fall (three salient excerpts below).

    Anthropogenic climate change is a theory whose basic mechanism is well understood, but whose magnitude is highly uncertain.

    The threat from global climate change does not seem to be an existential one on the time scale of the 21st century even in its most alarming incarnation.

    It seems more important that robust policy responses be formulated rather than to respond urgently with policies that may fail to address the problem and whose unintended consequences have not been adequately explored.

    These three comments are pretty straightforward.

    And I think they go to the heart of the matter in our discussion here of “black swans” (unknown unknowns of substantial impact) and their impact on climate science and policy.

    This is the real topic here, as I understand it.

    Max

    • Max, I had the lead letter in The Australian almost four years ago, making very similar points. Unfortunately, the politics of the issue have been all downhill since then, all efforts to seek a non-IPCC reassessment of the AGW data and case have been ignored, the government ploughs ahead with costly and ineffective measures. However, the tide has turned in public opinion. Where the AGW and need for urgent action was once widely accepted, the latest Newspoll show many more strongly against action than for it (though IMHO the questions asked were ridiculously badly framed).

  58. A poster here has equated “black swans” (Taleb) with “paradigm busters” (Kuhn?).

    Here is a potential one.

    The CLOUD experiment at CERN, Geneva, is currently investigating in a large-scale experiment the cosmic ray / cloud hypothesis, which Svensmark tested in a small-scale lab test.

    If this hypothesis is validated it could be a “paradigm buster”, in that it would provide a mechanism for the observed reaction of our climate to changes in cloud cover (Spencer) and the empirically observed correlation between solar activity and temperature (Svensmark), which lies beyond the measured impact of direct solar irradiance alone.

    Would it be a “black swan”? To some, maybe, but certainly not to Svensmark or CERN.

    Max

  59. Would this be a ‘black swan’ for AGW proponents?

    http://www.warwickhughes.com/icecore/

    • That’s definately a black swan, even for many sceptics.

      For me, it’s a white swan.

      • How many would regard this as a white swan?
        Not many I expect.

      • I’m fairly skeptical and I can’t help but be skeptical that burning fossil fuels wouldn’t increase the amount of co2 in the atmosphere. It is a black swan to me unless someone can explain why this wouldn’t happen.

      • Wow, I can think of dozens of ways, most of which boil down to the same mathematical principle: CO2 concentration is very possibily in a stable equilibrium. That is, there are dozens of reactions that might remove CO2 in a concentration-dependent fashion. For example, push CO2 concentration up a bit, and plants grow faster, sucking more out. Or more preciptates out into the oceans. Or, alternatively, reaction rates of things that normally put CO2 into the atmosphere go down.

        In fact, dynammic equilibrium is an almost universal rule in biolgical systems. If you push hard on a logistical curve, the result is just an oscillation about an asymptote with a decaying amplitude. (A soft push just gets you a monotonic approach.) That simple fact alone is probably the most important reason why my very first reaction to the CAGW theory was incredulity.

        So not only does it seem possible that it would not happen, it seems like the more reasonable presumption that it wouldn’t.

      • There may be dozens of chemical reactions that would influence the proportion of added co2 that stays in the atmosphere. It doesn’t change the fact that there is more of that element and it will seek a new level of equilibrium with a higher level in the atmosphere. The only way out of this would be if co2 was the limiting factor for a reaction in which case we would have seen decreasing levels of co2, is this not correct?

      • ps; before you answer with a biological reason why this may be possible think biomass trends and if the evidence supports your argument

    • Not really important, nor surprising. Such pieces are perennial as the grass, and get mowed down as routinely.

      Except and unless you accept the premise that all observations are completely useless because someone is able to quibble over their details, follow through the math, compare with the original works, and you still come up with much too little error to account for the 22%-25% differences seen over the past quarter millennium and the maximum CO2 level measured in the past 60K-85K years and extrapolated for the past 10-15 million years.

      Or you do; possibly, depending how willing to quibble away the best information available with the narrowest criticisms strung together to form a persuasively large but not altogether internally consistent objection.

      If you accept even the most skeptical interpretation of the most widely-reported ice core and related data and analyses without resorting to quibble, CO2 today is definitely rising; it’s definitely above the levels expected in the past 10-15 million years, it’s definitely rising faster than any quarter-millennium in the past 65K years, and probably rising faster than for any quarter-millennium period in the past 30 million years and it’s not entirely clear that even if all human contribution to CO2E levels were curtailed that the trend would stop (though it is abundantly clear that substantial human reduction in CO2E emission would slow the trend greatly).

      Or, unanswerable mountains of quibbles prevent any intelligent approach to any question at all.

      Before you ask, I also don’t call any claims that CO2 will save the world a black swan. More, like this claim, of a cuckoo’s egg.

    • No, they will say the author must be looking for oil industry funding.

      Maybe those black swans are just covered in oil!

  60. ‘We emphasize the importance of understanding dragon-kings as being often associated with a neighborhood of what can be called equivalently a phase transition, a bifurcation, a catastrophe (in the sense of René Thom), or a tipping point.’ Sornette 2005

    The term tipping point is a red flag for some – it has been used loosely to describe a cascade of processes leading to runaway global warming. The catastrophe of René Thom is of another order in which the processes are barely understood and the interactions not predictable at all in their dynamical complexity. I am of the opinion that cooling is more likely than warming simply because the world has been more frequently cooler in the past 2.38 million years.

    The bifurcation is a result of sensitive dependence – a trigger causing cascading changes in ice and snow, winds, clouds, dust, ocean circulation and vegetation. Tremendous energies cascading through powerful systems. Whatever the natural triggers are – it seems theoretically certain that anthropogenic greenhouse gas emissions is potentially a trigger of human origin. I will refrain from the obvious analogy to Russian roulette.

    The non-linear fluctuation of climate in the vicinity of a chaotic bifurcation is a Dragon King – such as occurred in Earth systems around 1910, the mid 1940’s, the late 1970′ and 1998/2001 – and not a black swan. Coincidentally the same temporal signature for changes in the trajectory for surface temperature? These Dragon Kings were fairly mild but there is a guarantee that they needn’t be.

    To me the policy path is obvious. We limit the great atmospheric experiment while building resilience in societies and economies because we will see another Dragon King sooner rather later. This is simply sweet reason and moderation – as opposed to the rank idiocies of both sides of the climate wars.

    • steven mosher

      “We limit the great atmospheric experiment while building resilience in societies and economies because we will see another Dragon King sooner rather later. This is simply sweet reason and moderation – as opposed to the rank idiocies of both sides of the climate wars.”

      yup

      • And “yup” again.

        Do we have a quorum yet?

        If so, we could move the debate to “how” to do that.

      • We build resilience in societies and economies every day already.

      • One would hope so. By resilience I mean food and energy security, democratic systems, economic development through free trade and good governance, health, education and sanitation. Good for population stabilisation and then the basics can be accomplished – ecological restoration and conservation, rebuilding carbon stores in agricultural soils, limiting black carbon and tropospheric ozone for health, environmental and agricultural benefit.

        Cheers Guys – I think we certainly have a quorum.

      • But now you have the same problem as with Pascal’s wager – how do you know that the dragon king that we will be dealing with is the dragon king that you’ve been expecting and planning for? How does Pascal know that Allah isn’t going to send him to hell for picking the wrong god?

      • Unlike Pascal, we don’t face a binary choice.

    • Products will be used when they become available. You cannot artificially increase the cost of an old product in order to encourage the introduction of a new product.
      (Especially when your have billions living in poverty)

      Artificially increasing the cost of an old product did not happen in any of the following:

      Gramophone=> Cassette player=> CD player=> iPod

      • Well, it is true that some people thought gramophones threatened the end of civilization as we know it; I confess that I’m not aware of anyone having made a technical argument that there is a “social cost of rock and roll” but metaphorically they certainly did!

        But pricing externalities is econ 101. And as any econ 101 student will tell you, you don’t want to address distribution (poverty) by failing to price social “bads”. If you believe that CO2 causes harm, the gramophone example fails. If you don’t, then the gramophone example is relevant.

      • Paul Baer

        CO2 + H2O + sunlight => plant food => animal food

        Therefore, CO2 is the foundation of life and extremely beneficial.

        Thanks to fossil fuels, life expectancy has increased steadily during its use. Thanks to fossil fuels, the naked animal (YOU) could live in the northern hemisphere and the Arabian desert.

        AGW is not supported by the data as shown in the following graph:
        http://bit.ly/iUqG8I

      • Girma,
        I don’t think anyone (and certainly not Paul) is saying CO2 in general isn’t beneficial nor that industrialisation hasn’t provided a million benefits. Rain is really useful but when it rains so heavily that the top soil of your farm is washed away, it is a problem for you. It is simply a matter of sustainability (which we might disagree about, but I think that is where we start).

        In the context of climate science, there is, in my view, no knockdown argument against one set of views or the other and that is why we are her debating the issues not debunking.

        http://mitigatingapathy.blogspot.com/

      • Well, of course you’re right that CO2 is essential for life, and there are certainly some measurable positive benefits of increasing it as well. But in general economists who have tried to sum the negative and positive benefits come up with large – potentially very large – net damages per ton of carbon.

        And I still don’t see how your graph of a steadily rising trend over 130 years does not support the theory of AGW. If not AGW, what explains the strong trend?

      • And I still don’t see how your graph of a steadily rising trend over 130 years does not support the theory of AGW. If not AGW, what explains the strong trend?

        The trend before wide use of fossil fuels about mid-20th century is identical to the trend after it, contradicting AGW theory: Increase CO2 => Increased rate of change of global temperature.

        As you said, the global mean temperature has a steadily rising trend for 130 years.
        http://bit.ly/iUqG8I

    • “…the rank idiocies of both sides of the climate wars.”

      This reminds me of Christopher Hitchen’s abandonment of the left in both foreign affairs and economics. He came to the conclusion that those with whom he had agreed most of his life (and with whom he had proudly manned the progressive barricades) were wrong on the most important issues of the day. Yet he could not bear to associate with those for whom he had had contempt his entire adult life (and who had for that same period held the positions he had by then adopted). So he decided to carve out a position from which he could disdain both sides of the arguments. His erstwhile progressive comrades were wrong now, but conservatives were also still wrong because they held the correct positions for the wrong reasons. Apparently, only two people in recent history where thereafter to be considered less than idiotic, Hitchens himself and, for some reason, George Orwell.

      “The people I disagree with are stoopid, and so are the people who agree with me,” can make for entertaining reading, but it is not a terribly coherent philosophy. But then no side of the debate (including the muddled middle) is immune to the human frailty of vanity.

      I would however love to hear what policies would “limit the great atmospheric experiment while building resilience in societies and economies.” This path may be obvious to some, but it would still be nice to see it outlined for the more simple of us.

      • You assume that one or other side is right – in human affairs I think that a rather bleak hope.

        I quote this from ‘The Wrong Trousers: Radically Rethinking Climate Policy, Gwyn Prins & Steve Rayner’ quite often.

        ‘Although it has failed to produce its intended impact nevertheless the Kyoto Protocol has performed an important role. That role has been allegorical. Kyoto has permitted different groups to tell different stories about themselves to themselves and to others, often in superficially scientific language. But, as we are increasingly coming to understand, it is often not questions about science that are at stake in these discussions. The culturally potent idiom of the dispassionate scientific narrative is being employed to fight culture wars over competing social and ethical values. Nor is that to be seen as a defect. Of course choices between competing values are not made by relying upon scientific knowledge alone. What is wrong is to pretend that they are.’

        So we have partisans in the climate wars – but which side are you on?

        There are some obvious ways forward in the Millennium Development goals, the Lomberg priorities. There is a practical outline in the Hartwell 2010 paper

      • I find that in human affairs one “side” in any important debate is frequently right, and the other wrong. Socialism vs. capitalism. Democracy vs. totalitarianism. Slavery vs. freedom. Rule of law vs. anarchy. Coke vs. Pepsi. And yes, consensus vs. skepticism.

        The debate most people are engaged in with respect to climate is whether CAGW justifies decarbonization of the national and global economy. Without that debate, and its enormous stakes both ways, there would be nothing but an interesting academic debate involving at most a couple thousand or so people. But here we are, with numerous high traffic climate blogs, vast amounts of research and advocacy dollars being spent, and elections globally turning in part on the issue.

        Have the climate scientists produced sufficient evidence of both causation and risk to justify massive changes in tax and regulatory policies so as to produce a carbon free environment? The CAGW/consensus/progressive position is yes. The skeptic/conservative position is no.

        Not to put too fine a point on it, when it comes to this central issue of the climate debate, there is no real middle ground. Think of it as a medical decision. One doctor says a leg must be amputated to save the patient. Another says there is insufficient evidence to justify such a drastic procedure. Where precisely is the middle ground? Either the CAGW advocates are correct, or they are not.

        There are certainly a host of other issues dealing with climate that can have different and varied positions, but not on the issue of whether CAGW justifies decarbonization. Diversification, black carbon, nuclear are all interesting and important issues, but that is not what all the fuss is about. When it comes to the main climate debate, you may certainly refer to both sides presenting “rank idiocies,” but don’t be surprised if some of us guilty of such supposed idiocy beg to differ.

        And no offense, but the statement that there are partisans in the climate wars is like saying there is water in the ocean. The Lomborgs and Pielke Jr.s are no different from the political”moderates” in any debate. I suspect the vast majority of “lukewarmers” would self identify as independent or moderate. Their proposals are simply not relevant to the central issue. Until that debate is settled, and it is not yet, their issues are unlikely to get much of a hearing. Once we decide whether to amputate or not, we can discuss the general health of the patient and what if anything should be done about it.

        As to which side I am on, well, if you are actually curious and not being rhetorical, read a comment or two of mine and I’ll bet you can guess.

    • Rob,
      You seem to overemphasize the role of chaotic bifurcations in the paper of Sornette. His dragon-kings are mostly due to largely non-chaotic processes. The growth of some cities is a good example. It seems to be natural that most countries have one city that is far larger than others. There are certainly self-enforcing mechanisms that have led to this situation. There might be something similar to chaotic bifurcation at some point of the development, but emphasizing those details is artificial and of little explanatory value.

      Similar self-enforcing mechanisms are behind many others of the dragon-kings. Sometimes they are of the nature of a tipping point, sometimes the case is that one contestant is bound to win and then winner the grow much stronger than others. Many processes are inherently unstable and bound to lead to one extreme, the only question remaining is the choice of the extreme (e.g., which of the early cities will grow to be the only really big one).

      The resulting situation is in most cases stable as long as external factors do not force a change, but attractors of a chaotic system may give the impression of such stability. This is, however, perhaps rather an exception than the rule. We should not expect to see chaos and attractors everywhere, when most cases are rather a combination of basic determinism and stochastic disturbances. This applies also to climate. The most likely description is determinism with stochastic influences, more fundamental chaotic nature cannot be excluded, but there are no good reasons to claim that it dominates. The combination of complex deterministic dynamics and stochastic disturbances may be similar to fundamental chaotic behavior over long periods, but there are differences. As an example the concept of equilibrium climate sensitivity is well defined without chaos, but not, when chaotic behavior dominates.

      • Pekka,

        Many systems have exhibit properties of dynamical complexity in theoretical physics. They range from nervous system organisation – in epilepsy for instance – to economic systems, numerical atmospheric and oceanic simulations, population dynamics and bushfire. They are known by their characteristic self organisation, sensitive dependence, slowing down and abrupt and non-linear change. The explanatory power derives from common properties in the meta theory of complex dynamical systems. It looks like a complex dynamical system that has numerous manifestations that we know of – as in the diverse examples of Sornette – therefore it has these properties that may make analysis tractable but in a different sense than we are used to.

        ‘We have presented supporting evidence for the concept that meaningful outliers (called “dragon-kings”) coexist with power laws in the distributions of event sizes under a broad range of conditions in a large variety of systems. These dragon-kings reveal the existence of mechanisms of self-organization that are not apparent otherwise from the distribution of their smaller siblings…’

        It is probably appropriate to consider city size as a dynamically complex system. ‘Actually, as is well known, Paris has played historically a crucial role in the development of France, and its dragon-king status observed here in the statistical distribution of French city sizes is a revealing sign of this rich and complex history. We will show in the following examples that the dragon-king status emerges in general from the existence of positive feedbacks, that amplify the role of certain events. In the case of Paris, the centralized organization for French governments over the past centuries has led to its ever-increasing pivotal role.’ But that particular example has no relevance to climate.

        There is a wealth of literature on non-linearity in climate dynamics. ‘The great revolution of nonlinear dynamics over recent decades has provided a wealth of information about the bifurcations that can destabilise a slowly evolving system like the Earth’s climate.

        These bifurcations are defined as points during the slow variation of a ‘control’ parameter at which a qualitative topological change of behaviour is observed in the multi-dimensional phase space of
        the system.’ http://arxiv.org/ftp/arxiv/papers/0907/0907.4290.pdf

        A good general summary of abrupt and non-linear climate change is provided in the NAS publication ‘Abrupt climate change: inevitable surprises’. Abrupt and non-linear climate change occurs on all timescales – and thus shows chaos as the underlying organising principle of climate.

        ‘The chaotic nature of the climate system was first recognized by Lorenz (1969, 1975), defining two types of problems associated with predictability:

        Predictability of the first kind, which is essentially the prediction of the future evolution of the atmosphere, given some knowledge of its initial state. Predictability of the first kind is therefore primarily an initial value problem, requiring a detailed set of good observations describing the actual conditions at the start of the modelling experiment. Daily numerical weather prediction is a typical example of this.

        Predictability of the second kind, in which the objective is to predict the evolution of the statistical properties of the climate system in response to changes in external forcings over time. Predictability of the second kind is essentially a boundary value problem, requiring good information on all external factors which might influence climate over time, e.g., variations in land use, ozone, aerosols, volcanic eruptions, solar variations, etc..

        Georgi (2005) demonstrates why climate prediction generally should be considered an initial value problem. To add difficulty to a prediction is the fact that the predictability of the climate system is strongly affected by non-linearities. A system that responds linearly to forcings is highly predictable, i.e. doubling of the forcing results in a doubling of the response. Non-linear behaviours are much less predictable and several factors increase the non-linearity of the climate system as a whole, thereby decreasing the predictability of climate systems in general. In addition to this, complex models involving nonlinearities and interactions tend to loose accuracy because their errors multiply.’ Professor Ole Humlum – http://www.climate4you.com/ClimateModels.htm

        Tsonis and colleagues identified chaotic climate shifts on regular intervals in the 20th Century – 1910, the mid 194o’s, the mid 1970’s and 1998/2001. These are apparent as abrupt changes in rainfall regimes globally and in the ENSO. The 1997/1998 El Nino would be a Dragon King for instance.

        Very much to the contrary Pekka – dynamical complexity is most obviously the underlying mode of operation of climate. Simple determinism is limited to the slow variation of ‘control variables’ and there is very little room for randomness at all.

        Cheers

      • Rob,

        All the observations that you present are plausible narratives that are produced after the observation. Separate processes that have afterwards observed to have similarities do indeed have those similarities, but those posterior observations do not mean that we could take advantage from these similarities in forecasting are making other deductive statements about the future.

        The space of all possible future behaviors is infinite. Similarities of two disparate processes in the past do not lead to valid expectations of similar futures, unless we really understand the dynamics of both processes and can deduce from that knowledge that the futures will be similar.

        Looking at a wide variety of different processes may help imagination, it may help in finding real dynamical regularities, but our ability to foresee something about future is improved only through this indirect process of learning, not from the historical similarities alone.

        Concerning the climate processes I have not seen any convincing evidence on real bifurcations or knowledge on dynamics that would lead to bifurcations. As in most application of chaos theory, we see only some past behavior, which can be linked to mode shifts or transitions from one attractor to another by people who want to present such interpretations. There is, however, no evidence on the correctness of these interpretations and they are anyway of little value as long as we cannot describe the dynamics well enough to study the onset of chaos and to classify typical attractors.

        For a complex dynamical system we have the possibility that dissipative processes are strong enough to maintain long term stability and to remove the possibility major state shifts. Such a well behaved system may still allow for large fluctuations of the type we know to exist in the Earth system, but such a system may allow for successful modeling and may make the climate sensitivity a well defined and important part of describing the reaction to increased CO2.

      • Pekka,

        The aspects of abrupt change in climate occur at all scales. ENSO we know is a multi-dimensional phenomenon involving upwelling, wind, cloud, Rossby and Kelvin waves. It is one of a broad number of Earth sub-systems that exhibit abrupt and non-linear change and broadly bi-stable behaviour. The AO, NAO, AMO, SAM, IOD, PDO, PNA etc. It seems more appropriate to define these as standing waves in a spatio-temporal chaotic Earth system than as random variability. Here for instance is an 11,000 years ENSO analysis – http://cat.inist.fr/?aModele=afficheN&cpsidt=2210907 – but there is a great deal of other literature available for perusal.

        You make a number of assumptions about climate behaviour that well-behaved or strongly dissipative that are not merely simply statements of what might be the case but seem clearly contradicted by the existence of abrupt and non-linear change in the system. These are the conditions that suggest bifurcation in a complex dynamical system.

        To me it patently absurd that the interaction of component parts of the Earth system evolving both temporally and spatially should not be defined as a chaotic system – as has been discussed for decades now and confirmed for instance in the recent Royal Society climate summary. Internal climate variation was said to occur because climate is an example of a chaotic system. You seem well behind the game in both theory and understanding of the data and present mere suppositions and possibilities. This is typical of someone defending a position but not of an open mind.

        In a truly chaotic system – as in the atmosphere and ocean simulations as well – prediction is problematical. It is in theory deterministic but in practice shows that we have some way to go to understand the details.

        Tim Palmer – whom I quote frequently – says that climate can only be predicted as a probability density function. The other paper on noisy bifurcation – noisy being another term for Dragon King – I linked to suggests that we may be able to predict tipping points by looking at slowing down and Dragon Kings in particular. This implies that predictability is not possible beyond the next bifurcation.

      • Rob,

        You continue to make strong claims on issues that are not proven – and you talk about open mind.

  61. Judith – you give two instances of Black Swans in climate – recent severe NH winters and the Queensland floods. For that to be so you would have to show that each event was both unprecedented and universally believed to be impossible, and that the events were ONLY explained in hindsight. But that isn’t true (how many times to we have to go over this?), is it? In each case, meteorologists, climatologists, hydrologists and engineers recorded their beliefs that these events were not merely possible, but likely. Climate “science”, of course, did believe these events were so unlikely that public policy ought to treat them as impossible. Climate “science” was wrong, and anyone following the pathetic defence put up on its behalf here and elsewhere by its dwindling band of believers can see why it was wrong.

    To treat these events as Black Swans, therefore, is to repose continued trust in the opinion of a group of people who ought long ago to have forfeit it, simply because they were largely successful in shouting down dissent. And just by the way, it’s also extremely disrespectful of those dissenters who through patient (if derided) application of good scientific method (as distinct from Climate “Science”) were able to make sound forecasts which, had they been heeded, would have saved a lot of pain and suffering. Instead of piling insult on injury, isn’t it time you started doing the opposite, in view of what you now know about the way your discredited craft has treated these people over the last decade or so?

    The winters and the floods were not Black Swans, they were straightforward disconfirmations of AGW theory, and deserve to be treated as such. It’s depressing to see such credulousness so long after Climategate. It seems that even you don’t yet understand the depth of the mischief your field has perpetrated.

    • “Climate “science”, of course, did believe these events were so unlikely that public policy ought to treat them as impossible.”

      I believe that this is absolutely false. But if you can demonstrate citations or quotes in which climate scientists said severe winters or queensland floods should be treated as impossible, I will admit my error. Quotations from the IPCC are especially welcome as evidence.

      • The premise of my comment was that JC’s characterisation of the severe winters and the Queensland floods as Black Swans. I believe they were not, because plenty of people thought they were far from unlikely, and said so BEFORE the event. You seem to be saying Climate “Scientists” should be included in this number. If you have any evidence for this I’m sure the Queensland Coroner would be interested to hear from you, as I imagine he is taking a keen interest in what advice was available to managers of public services, and from whom.

        And while you are at it you’d better explain to Judith that, far from being Black Swan events, severe NH winters and Western Pacific floods were in fact all scenarios warned of by Climate “Science”.

      • It was covered in the part where the IPCC said just trust us, bad scary things will happen if you stop emitting all your co2.

      • I’m not sure how severe NH winters can be classified as scenarios warned of by climate science. You should have fewer cold extremes. You should have first snows later and last snows earlier. The snow cover should decrease both due to some areas getting no snow and others getting snow later and having it melt earlier. Perhaps you have a definition of severe NH winters that would fit in with projections that hasn’t occurred to me? I have to tell you that you would be much better off yelling weather then you would be yelling AGW when looking at snowstorms in places like Alabama and Georgia. The recent snow extent coverage for March is also not conducive to your argument in any imaginable way.

      • Are cold winters in Europe associated with low solar activity?
        M Lockwood1,2, R G Harrison1, T Woollings1 and S K Solanki3,4

        http://iopscience.iop.org/1748-9326/5/2/024001

        If Lockwood et al are right about solar UV – I’m pretty sure nearly everyone missed it.

        I am a Queensland hydrologist. We have periodicities of 20 to 40 years of intense and frequent La Niña followed by a intense and frequent El Niño. We are in a cool La Niña period and the current super La Niña (and resultant flooding) is a result of that. The decadal variations have an impact on global surface temperature as well. I have looked and looked – and this is not something that the IPCC picked up on at all.

        Is this an example of the strong confirmation evident in science in the article Girma referenced elsewhere? Whatever it is – the track record of the IPCC is appalling.

      • In addition to the PDO there is also the AMO to take into consideration. The studies of this ocillation all are very similar in regional influences on temperature. If you look at the maps of these influences it doesn’t take make imagination to picture what the responses will be to a switching from the positive to the negative phase.
        http://www.meteo.psu.edu/~mann/shared/articles/KnightetalGRL05.pdf

      • I see both my comments are empirical evidence that commenting when woken up in the middle of the night may lead to errors.

      • No, I did not say that climate scientists were among those warning that severe Queensland floods or cold NH winters should be planned for. Rather I’m rebutting the claim that climate scientists “did believe these events were so unlikely that public policy ought to treat them as impossible.” That’s a positive assertion that requires evidence, for which I doubt there is any.

      • John Kannarr

        I suggest that the more typical situation was that sensational nonsense appeared in the MSM by MSM reporters and environmentalist extremists, such as the claim that soon schoolchildren wouldn’t know what snow was like. The climate scientists certainly didn’t, as far as I can tell, rush to disabuse the MSM of such notions or to contradict them. The effect was to let the public be deluded about such things, by those who hoped that the public would rise up and demand politicial action, while the climate scientists could comfortably sit back, let the wild claims appear to be part of their famously “settled” science, knowing that if the “predictions” failed, they could point to their refereed journal papers that made no such explicit claims, or at least none with claimed certainty, thus achieving sensational scare stories but with plausible deniability.

        Nice little con game.

      • John Kannarr

        Sorry, my comment was meant to be a response to Paul Baer’s response to TomFP in which he asked for “citations or quotes in which climate scientists said severe winters or queensland floods should be treated as impossible.” Instead, the climate scientists let the MSM carry their water for ridiculous claims, while retaining plausible deniability.

      • John Kannarr

        I suggest that the more typical situation was that sensational nonsense appeared in the MSM by MSM reporters and environmentalist extremists, such as the claim that soon schoolchildren wouldn’t know what snow was like. The climate scientists certainly didn’t, as far as I can tell,

        Actually it is a scientist who said that!

        Snowfalls are now just a thing of the past
        20-March-2000
        http://ind.pn/i2ZHaw


        Global warming, the heating of the atmosphere by increased amounts of industrial gases, is now accepted as a reality by the international community. Average temperatures in Britain were nearly 0.6°C higher in the Nineties than in 1960-90, and it is estimated that they will increase by 0.2C every decade over the coming century. Eight of the 10 hottest years on record occurred in the Nineties.

        However, the warming is so far manifesting itself more in winters which are less cold than in much hotter summers. According to Dr David Viner, a senior research scientist at the climatic research unit (CRU) of the University of East Anglia,within a few years winter snowfall will become “a very rare and exciting event”.

        “Children just aren’t going to know what snow is,” he said.

      • Girma

        People (including venerable scientists) often make totally stupid predictions, as you pointed out to John Kannarr

        Here is one of the most absurd:
        http://www.independent.co.uk/environment/why-antarctica-will-soon-be-the-ionlyi-place-to-live–literally-561947.html

        Antarctica is likely to be the world’s only habitable continent by the end of this century if global warming remains unchecked, the Government’s chief scientist, Professor Sir David King, said last week.
        He said the Earth was entering the “first hot period” for 60 million years, when there was no ice on the planet and “the rest of the globe could not sustain human life”. The warning – one of the starkest delivered by a top scientist – comes as ministers decide next week whether to weaken measures to cut the pollution that causes climate change…

        Scientists are not immune from spreading foolish BS, it seems.

        Max

      • The tone of the statement was certainly alarmist by purpose, but the most extreme conclusions were obvious misinterpretations by the journalist. What King really said according to the news was:

        – CO2 concentration will reach 1000 ppm by 2100.
        – That level will ultimately lead to an iceless Earth.
        – The iceless Earth will be too hot for humans except on Antarctica.

        It’s clear that he did not say that ice would disappear by 2100. Neither did he claim that the Antarctica would be the best place to live by 2100.

        This is really a prime example of misquotes by journalists.

      • Actually King didn’t say even as much as I wrote above. He did not make other statements about future than that the CO2 concentration will reach 1000 ppm by 2100 if nothing is done to reduce emissions.

        All the other comments ware about the situation 60 million years ago. He clearly hinted that something similar may result from the high CO2 concentration, but the article does not tell that he would have claimed that this is a likely development.

      • Pekka Pirilä

        Here is what Professor David King actually said:
        http://www.ofcomswindlecomplaint.net/Misreprestn_Views/davidkingviews.htm

        Fifty-five million years ago was a time when there was no ice on the earth; the Antarctic was the most habitable place for mammals, because it was the coolest place, and the rest of the earth was rather inhabitable because it was so hot. It is estimated that it was roughly 1,000 parts per million then, and the important thing is that if we carry on business as usual we will hit 1,000 parts per million around the end of this century..

        Hooo!

        1,000 ppmv CO2 caused the Earth to be inhabitable for mammals (except for Antarctica) 55 million years ago?

        And “if we carry on business as usual we will hit 1,000 parts per million around the end of this century”…

        Ouch! How could a scientist allude at something so utterly absurd?

        Max

      • No Pekka, it is NOT a simple case of misquoting.

        Read what King actually admits to having said.

        That’s absurd enough, without requiring a misquote.

        Max

      • It was serious misquoting.

        King may have said stupid things, but he didn’t say, what the news article claimed.

      • Pekka Pirilä

        You are truly bending over backwards to defend Professor Sir David King.

        What he said was totally absurd (that Antarctica would EVER be the only habitable part of the globe – for humans, not penguins).

        If the reporter’s quote was even more exaggerated by putting a 2100 time-line on it, that’s quite normal for reporters.

        And, if this were the case, why did King not subsequently go on record to correct the reporter rather than let the absurd statement stand?

        But the point of the matter is that a venerable scientist made an absolutely absurd public remark, no matter how much this may offend your sensibilities.

        But it is no more absurd than another scientist’s rant about “coal death trains”.

        Moral of the story: Scientists are human. To err (at times) is human. Ego, scientists err (at times) like anyone else.

        Sometimes they even exaggerate to the point of absurdity (as happened in the case in point).

        Max

      • This was a simple case of misquoting. Wrong evidence is not evidence.

      • Paging Dr. Danny Bloom, STAT!
        =============

      • Is there any credible evidence at all that Earth’s atmosphere will have a CO2 concentration of 1000 ppm by 2100?
        No.
        This was alarmist crap by Dr. King and now the true beleivers are busy performing apologia for him.

      • Too bad! The planet’s flora have eaten themselves into near CO2-starvation, and it’s up to we fauna to pick up our game! I suggest we target 2,000 ppm.

      • oneuniverse

        Hi Pekka,

        By not mentioning the time required (thousands of years according to King) for all the ice to melt, were it to melt, Prof. King left the journalists to draw their own conclusions (“Prof. King said there was no ice when it was 1000ppm in past, so it sounds like there’ll be no ice if 1000ppm in 2100” ).

        Neither did he mention the timeframe in his interview in the Observer.

        Neither King nor the Climate Group (at whose launch he was speaking) corrected the gross mistake when it was printed in the mainstream UK press.

        It strikes me as irresponsible to provide the press with partial information with obvious potential for misunderstanding, and to then not correct the misunderstanding when it duly happened.

      • maksimovich

        This one from Nasa is a fine example

        Continued operation of the oceanic conveyor belt is important to northern Europe’s moderate climate because of northward transport of heat in the Gulf Stream and North Atlantic Current. The system can weaken or shut down entirely if the North Atlantic surface-water salinity somehow drops too low to allow the formation of deep-ocean water masses. This apparently happened during the Little Ice Age (about 1400 to 1850 AD). The conveyer system shut down and northern Europe’s climate became markedly colder. Old paintings from this era show Dutch skaters on frozen canals-something that would not occur during today’s climatic regime. Cores extracted from deep-sea sediment deposits contain evidence of earlier cold periods.

        http://oceanmotion.org/html/background/ocean-conveyor-belt.htm

        Ah well

        http://www.jaunted.com/story/2009/1/10/132848/432/travel/Pass+the+Dutchie%3A+Queen+Beatrix+Laces+Up+Her+Ice+Skates+as+Canals+Freeze+in+Holland

      • Come on – its not just climate scientists who get things wrong in some area of expertise – who can forget:

        Spam will be a thing of the past in two years’ time

        http://news.bbc.co.uk/1/hi/business/3426367.stm

      • Max

        Here is another interesting quote by NASA (NASA Facts April 1998, NF-222) when science used to be only about the truth:

        … For example, in the early 1970’s, because temperatures had been decreasing for about 25 to 30 years, people began predicting the approach of an ice age! For the last 15 to 20 years, we have been seeing a fairly steady rise in temperatures, giving some assurance that we are now in a global warming phase.

        http://bit.ly/ehUDkB

        Unfortunately, you cannot find that article at NASA’s website now.

        Sad.

      • Although if you read down to the end of the article, he also said this:

        “Heavy snow will return occasionally, says Dr Viner, but when it does we will be unprepared. “

      • John Kannarr

        Okay, I may have been too optimistic, or giving the benefit of the doubt, in claiming that scientists do not make many of the sensationalist claims, but allow journalists to do so as though the claims were from the scientists.

        I see that the British Government’s chief scientist, Professor Sir David King is quoted in the article cited at http://www.independent.co.uk/environment/why-antarctica-will-soon-be-the-ionlyi-place-to-live–literally-561947.html as saying: “No ice was left on Earth. Antarctica was the best place for mammals to live, and the rest of the world would not sustain human life.”
        I doubt that many reading the article would be aware that due to plate tectonic movements, 60 million years ago, what is now Antartica was at roughly 60 degrees South latitude (and certainly some parts of what is now Northern Asia were at least as far as 60 degrees North, so would have presumably been as habitable as “Antarctica”), and very likely we have no definite knowledge of the situation regarding sea ice at either pole at that time. A little bit of information can go a long way in misleading people in the hands of a journalist advocate, or of a scientist advocate.

      • Paul Baer

        I believe that this is absolutely false. But if you can demonstrate citations or quotes in which climate scientists said severe winters or queensland floods should be treated as impossible

        Leading environmentalist Professor Tim Flannery has warned that Australia is now entering long-term climate change, which could cause longer and more frequent droughts.

        He also predicts that the ongoing drought could leave Sydney’s dams dry in just two years.

        http://bit.ly/jif1SK

        The above statement was made by the professor on June 11, 2005, six years ago.

        Six years ago, Sydney dam level was at 39%: http://bit.ly/kTLLki

        Now, six years latter, Sydney’s dam is not dry, but it is 74.3% full: http://bit.ly/jsbgNy

        Professor Flannery, do you accept now that your prediction of dam level was wrong?

        Why do they scare monger?

      • could

        Lol.

    • TomFP

      The winters and the floods were not Black Swans, they were straightforward disconfirmations of AGW theory, and deserve to be treated as such.

      Here is reinforcement for your conclusion.

      Study of global mean temperature since 1880 of about –0.2 deg C shows a 30 years global cooling of 0.4 deg C followed by a 30-years global warming of 0.7 deg C.
      http://bit.ly/iUqG8I

      Verification:

      1880 to 1910 cooling => -0.2-0.4= -0.6 deg C

      1910 to 1940 warming => -0.6+0.7= 0.1 deg C

      1940 to 1970 cooling => 0.1-0.4= -0.3 deg C

      1970 to 2000 warming => -0.3+0.7= 0.4 deg C

      Such a single pattern existing before and after mid-20th century contradicts AGW.

  62. “Why, practically every nonsense that has ever been said about capitalism has been championed by some professed economist”
    Joseph Schumpeter

    The following is taken from an article by Sinclair Davidson called ‘Economists as Social Engineers’ . There a great divide in the professions of economics. Some believe in intervention in the vein of Keynes and Pigou. Some find a more convincing lineage in Hayek and Schumpeter. The frequency of black swans in our economies suggest that Hayek had a prima facie case for the uncertain fate of interventions in the market.

    The polls suggest about a 50/50 breakup between pro and anti carbon tax positions in economists. Today’s poll in Australia shows a general 30% support for a carbon tax. It is dead and buried in America. We know that most nations declared a no tax position at Copenhagen. This should be your first practical position Paul – the pointlessness of flogging a dead horse.

    We would prefer that neither you, Bart or Martha lecture us on the virtues of the free market. “The results speak for themselves. The market economy needs no apologists and propagandists.” The human condition has improved dramatically since 1776 when Adam Smith published his magnum opus. For a free market to survive, its friends need to speak up. Capitalism and market forces have generated sufficient wealth to provide a higher education for an ever-increasing proportion of the population. Mises tells us “what determines the course of a nation’s economic policies is always the economic ideas held by public opinion. No government, whether democratic or dictatorial, can free itself from the sway of the generally accepted ideology.”

    Adam Smith did not start that process, and indeed had nothing to say about the early industrial revolution that was occurring around him. He described the process whereby human wealth has increased. That understanding, however, has not survived in modern economic theory and is not taught in our schools and universities. Indeed, the results do not speak for themselves. Both Hayek and Schumpeter explain why.’

    I think that Sinclair Davidson – as Professor of Economics at RMIT – is not entirely correct about modern economic theory in our schools.

    ‘Little else is requisite to carry a state to the highest degree of opulence from the lowest barbarism, but peace, easy taxes, and a tolerable administration of justice; all the rest being brought about by the natural course of things. All governments which thwart this natural course, which force things into another channel or which endeavour to arrest the progress of society at a particular point, are unnatural, and to support themselves are obliged to be oppressive and tyrannical.’
    Asam Smith

    The pricing of externalities is in fact very difficult as every student of environmental economics can tell you. Truth to tell though you are not concerned with the pricing of Pigouvian externalities but of increasing the cost of energy to that whereby it is supplanted by another technology. A different matter entirely. That this is guaranteed to place massive pressures on costs throughout the developing world is reported by Shi-Ling Hsu.

    ‘Again, climate scientists should not be any more surprised by the apathy than they are by the violence. If climate scientists are right, then the world faces a stark choice: either undertake fundamental changes in the way that almost every economy operates, imposing substantial costs on almost every country and society in the world, or roll the dice and see what happens with the Earth’s climate.’

    The cost seem potentially genocidal as Girma assserts. I am fairly certain – if these are the only choices on the table – that the world will roll the dice. Your troubles are cumulative if the world cools for another decade or 3 – as is an emerging theme in peer reviewed literature. Again, we get back to the impasse – and it is one that we can only get past by surrendering carbon taxes as an option. What we need to do this century is increase food supplies and energy by 3% a year – not doing this is not an option.

    There are alternatives that I have discussed with you before I believe. Very briefly they involve reduction in black carbon and atmospheric ozone. Very effective and good for health, the environment and agriculture. Access to health care, education, safe water and sanitation would do more than anything to stabilise population. Conservation and restoration of ecosystems and agricultural soils has the potential to sequester vast amounts of carbon and provides food security and biodiversity benefits.

    I have really just touched the surface here in what could be done. It simply needs to world to focus of the achievable and the pragmatic. But taxes? You may as well not bother because the argument has already been lost and we need to move on. But if all you are interested in is scoring points as a climate warrior on easy targets – by all means continue.

    • Chief

      Who micturated on your Corn Flakes?

      Popularity contests and polls from a scientist? An engineer, no less? A radical contrarian who would rather be right than be popular and would rather go to jail than vote?

      I get bipolarity, Chief. Really, I do. But this post of yours is as Black Swan as it gets, coming from you and considering your past record of intellectual accomplishments.

      Well, not the part about utterly failing to grasp Economics, that’s pretty consistent, and to be expected from someone firmly rooted in the hard sciences, however much progress you’ve made in understanding Chaos and complex systems.

      I get that the proposed Australian carbon tax goes half to some bumptious government programs and only half to the shareholders of CO2E, and is set at an absurdly low arbitrary level with no real plan for right-pricing this common asset, so is bound to satisfy no one.

      But what I’m not grasping is this utter surrender to incompetence that you embrace here, Chief. When did pandering to the lowest common denominator become your modus op, in place of fighting the good fight?

    • “The frequency of black swans in our economies suggest that Hayek had a prima facie case for the uncertain fate of interventions in the market. ”

      Umm, it seems like the long history of bubbles, long waves, etc. long before there was anything like modern fiscal and monetary policy suggests that Hayek’s prima facie case is not a case against intervention per se.

      • There are some essential governance issues – and these include oversight of financial prudence and management of interest rates to avoid speculative bubbles. Both were neglected in the US prior to the GFC leading to an outcome that Hayek predicted in the 1930’s – so hardly a black swan.

        This is not a case of being right or wrong – but I would put Democratic Socialism into the realm of the idiotic.

    • Chief

      “The cost seem potentially genocidal as Girma assserts. I am fairly certain – if these are the only choices on the table – that the world will roll the dice. Your troubles are cumulative if the world cools for another decade or 3 – as is an emerging theme in peer reviewed literature.”

      Okay, on costs, as it comes up repeatedly.

      1. Asserting that costs will increase is just that, an assertion, and no more. There is and can be no rational basis for the claim.

      Are some forms of energy more costly than others?

      Certainly.

      However, the comparison in the marketplace isn’t between the utility of one form of energy and another; it is between the utility of one good and another.

      Given the choice between an 11 mpg and a 61 mpg vehicle – all other things held to be equal from the point of view of the utility of the vehicle to the individual buyer – it seems ridiculous that anyone would elect 11 mpg.

      [Rant starting]However, this was the exact choice made in America 15 years ago. Why? The actual reason was a concerted marketing campaign by the auto manufacturers to steer buyers toward the 11 mpg vehicle due to the immense profit margins of the lower-cost, lower-manufacturing-standards, heavily subsidized 11 mpg units. There was a misleading ‘safety blitz’ using scare tactics targetted to frighten female buyers (and relentlessly sharpened through focus groups to produce exactly that effect) when in fact the 11 mpg vehicles were less safe and produced more dangerous conditions overall. There was a taste-shifting campaign through product placement in the popular media to influence younger drivers. There were strong and successful lobbying campaigns to not merely subsidize part of the price of the 11 mpg vehicles, but in some cases governments ended up paying more per 11 mpg vehicle than the price of the vehicle in the showroom altogether between federal and state incentive programs, before counting the cost of paving the roads or infrastructure.[End rant]

      There is no reason that America ought not have been consuming 80% less fuel on the roads for the past decade or more, other than the profit of a few obtained by subsidies and chicanery.

      With 80% lower demand for gasoline in America, can you imagine how much lower the price of fuel, how much more utility from the market individual consumers would have obtained, how much better off the general populace of America would have been?

      How much less insanity in oil-producing regions would have dominated?

      What experience tells us is that some stakeholders cannot be trusted to do what is right for all stakeholders, and in these cases the power to do wrong ought be curtailed for the sake of the fair market.

      Less state-sponsored corporate socialism, less hand-holding of companies who in their dotage crush the opportunities of new enterprise while strangling the democratic decision power of their customers and suppliers, no tolerance of anti-American behavior by business.

      Smaller ‘government’, if you can call people who dig into the tax stream up to their armpits and give cash to failed business models hand over fist ‘government’.

      Now, how do you call what actually happened with cars in America since the SUV, ‘lower energy cost’?

      In other areas, sure there are switching costs to overcome; however these are largely one-time, small charges that once disposed of lead to lower energy costs too.

      Switch from inefficient appliances to efficient ones, or in the case of appliances sitting unused or suboptimally used, get rid of them and replace them with other options.

      How many Americans use refridgerators to hold nothing more than beer?

      How many use freezers only to make ice and hide valuables?

      Replace most heaters with heat pumps, and you save over 70% energy costs; more, if air conditioners are also used in the same location.

      With an extra $1000 a year income, how many people wouldn’t move to within walking distance of all their needs, in a swankier neighborhood, and save themselves both the expense of gasoline and car (and insurance, and parking, and maintenance and repairs)? Would they consider themselves worse off, or better off, living in a better home just because they use less energy?

      And what of the ‘human dignity’ question? How can the rest of the world improve its lot if not by burning more fossil fuel? Same way as the developed world would improve its lot by burning smarter, not harder.

      The ‘cost more’ argument is invalid and defunct. Move on.

      2. The cooling world. You must know better than to trust your own predictions so completely, Chief.

      Not that they’re wrong, or right. But a world cooling for 30, or 70 years, need not mean anything about the +/-AGW debate. Random claims about temperature mean nothing about the real issue, which is the shaky artificial scaffold of CO2E being built higher and higher in the climate, and what happens when that scaffold tumbles down, or what natural flows that artificial structure will one day perturb so much as to extinguish something like the appearance or disappearance of an intermittent gyre like ENSO.

      Girma’s wager that future data noise will overwhelm current data signal is a waste of my time, and yours. Why go there?

      • Bart R

        Girma’s wager that future data noise will overwhelm current data signal is a waste of my time, and yours. Why go there?

        Prediction in 2000:

        it is estimated that they will increase by 0.2C every decade over the coming century.

        http://ind.pn/i2ZHaw

        Observation in 2011:

        Global cooling: http://bit.ly/fWxIYn

      • Girma

        As silly as the 0.2C claim may be, one notes that 11 years is not a century; I’m unwilling to wait to dismiss a bad method based on outcomes, when I can dismiss bad methods based on known methodological failings.

        Indeed, outcomes do not always invalidate the mechanism underlying predictions.

        It took the better part of a decade for the first experimental results to confirm the mechanisms Einstein set out in his five papers in 1905; with many experimental outcomes apparently discomfirming his predictions.

        The prediction-wager is an old fallacy.

        It ought not be engaged in; where engaged in, not endorsed.

      • steven mosher

        That’s a fun game.

        1. Since I believe in sensitivities under 3, I wouldnt buy into
        the .2c warming
        2. The way to test a prediction is well known.

        you dont start 2002. thats not the start date of the prediction.

        here:

        http://www.woodfortrees.org/plot/hadcrut3vgl/from:2000/to:2011/plot/hadcrut3vgl/from:2000/to:2011/trend

        global warming.

        fun game. next.

      • Technically, the prediction originated in FAR, in 1990, with the IPCC:

        http://www.woodfortrees.org/plot/hadcrut3vgl/from:1990/plot/hadcrut3vgl/from:1990/trend

        Though again, it’s silly and I’m not endorsing this clearly benighted and pointless exercise.

      • Steven mosher

        In your graph, what is the global warming rate since 2000?

        only 0.05 deg C per decade

        Four times the predicted value of 0.2 deg C per decade.

        They exaggerate by 4-times.

        http://bit.ly/jpDMXj

      • Sorry

        Steven mosher

        In your graph, what is the global warming rate since 2000?

        only 0.05 deg C per decade

        [Five] times the predicted value of 0.2 deg C per decade.

        They exaggerate by [5]-times.

        http://bit.ly/jpDMXj

      • Uh, Girma –
        [Five] times the predicted value of 0.2 deg C per decade.

        They exaggerate by [5]-times.

        0.05/0.2 = 0.25 = One quarter the predicted value
        They exaggerate by [4]-times.

        Take your time, check your numbers, don’t let the termites eat your attitude. :-)

      • Thank you Jim.

        Sorry.

      • Or, since we’re playing this game, if we went by the definitions used in FAR, based on 30-year trends (the standard at the time), there’s been roughly a 0.5C increase between 1981 and 1011; 0.5C/3 decade = 0.17C, a mere rounding error from 0.2C/decade (though only 1/3rd of the top of the IPCC FAR range of 0.5C/decade).

        So, based on the conditions set out when the game started, IPCC is within striking distance of a correct prediction, while Girma’s characterisation of the score appears decidedly one-sided.

        Not that it makes a difference.

        0.2C/decade is a silly premise.

      • Bart –
        roughly a 0.5C increase between 1981 and 1011

        1011? Really?? :-)

      • steven mosher

        actually you have to look at the error bounds of all the elements.

        1. you have the models. the models are averaged into a model mean.. .2C. that mean has an error bound. Guess what, in the short term that error bound is large in the long run it gets narrower.

        2. you have the observations. When you “fit” it with a straightline you are making an assumption and an estimate.
        The assumption is that a straightline model is good to capture the trend. that assumption has uncertainty. when you calculate the trend you also have uncertainty.
        There are several ways to adress this. You have availed yourself of NONE of the standard statistical techniques are tests.

        You do your cause no favor by engaging in statistical analysis that is worse than the practices some of us have criticized in certain climate science work.

        If you want to see how this is done right

        Go here.

        http://rankexploits.com/musings/2011/hadley-march-anomaly-0-318c-up/

        If you want to learn how to compare the trend of the model means against the trends in the observations, go to that site.
        we’ve been looking at it since 2008 or so.

        I believe Lucia and a couple others have a paper in the process showing the methods. my guess is you dont have the guts to join a discussion on the issue.

        Lucia will answer all your questions.

      • stephen mosher

        I’m not entirely sure what to say here.

        Lucia’s methodology is persuasively nuanced, though one does have questions.

        One notes Lucia admirably prefers to look at predictions based on data available after the prediction was fixed.

        However, Lucia then chooses — I’m sure for amply valid reasons — 2001 as the date when the prediction set out in 1990 was fixed.

        Seems arbitrary to me based on the little information I have; though again, I’m certain Lucia and I are likely discussing different things.

        I’m going from the first time IPCC made a 0.2C/decade prediction, and I’m certain Lucia is using some later event.

        Also, this whole multi-model mean nonsense..

        There are many advantages of looking at multiple models, if you have enough of them and they have some validity.

        Deriving a sound basis for prediction based on their mean is not one of them, I think, unless the models agree with this position, and the models don’t seem to indicate prediction is possible, due sensitivity to initial conditions.

        Until models start predicting unpredictables like volcano activity, forest fires, and all of those other factors that tend to put significant pressure on the temperature trend, I’m thinking it’s a waste of time to play statistical hide-and-seek on them over the short run.

        Indeed, in the long run, there’s a good chance that the chaos in the system will, perturbed by these unpredictables, overwhelm many mechanisms that currently play a role in climate — AGW, if it is a mechanism in climate, could be one such overwhelmed effect.

        However, the question remains, is it wise to continue to perturb that system more with more CO2E emission, given that we know its chaotic nature?

      • steven mosher

        Bart R.

        Youre confused. go there and see how well your arguments survive. But they will have to be math arguments.

        “2001 as the date when the prediction set out in 1990 was fixed.”

        The question is Ar4. For that prediction we have the relevant model data to create an estimate and check it.
        In my mind previous predictions are OBE. The same with Ar4 when Ar5 comes out. Ar5, however, will be better since they will do decadal predictions for sequence of decades.

        as for volcanos you will have to wait for Ar5 where you get a variety of sensitivity runs on that. Since you cant predict a volcano you are limited to exploring the uncertainty.
        1. what happens if no volcanoes.
        2. what happens if you get an “average” number
        3. what happens if…

        2 and 3 are not policy relevant.
        same with forest fires or any other ‘unpredictable’

      • steven mosher

        You sure know how to make a feller feel welcome.

        Call him gutless and invite him to put his arguments to a survival test at the hospitality of another?

        Seeing as I find the entire exercise of applying even the best statistical tools to a statistically misbegotten graph a complete waste of time, I’m comfortable with declining your kind offer on Lucia’s behalf.

        Please extend my sincere regrets.

        It’s clear the models from 2001 handle poorly or not at all ENSO — hardly surprising given how sparsely represented ENSO was up to the late 1990’s.

        If they somehow get good at volcanoes and forest fires, it’s still insanely ill-advised to take their mean and call it a prediction, unless the models somehow overcome the myriad issues of sensitivity to initial conditions and themselves agree that models ought be predictive. Do they do that now? Do you expect them to do that soon?

        I’ve seen only limited suggestion in the literature that such is even plausible, and I remain dubious this will change much.

        And by the time such models produce much, in that distant future, one can be certain their error bars will be so wide as it will take many decades to decide whether the hypothesis of any of the models, much less their mean, can be rejected.

        Or, one could test Lucia’s method for rejecting the models as of 2001 by hindcasting, and applying it to running start years before 2001.

        How many times does the actual data reject the hindcast model on a decadal trend line?

        I imagine it’s a significant fraction of the time, even though the models were trained on that same data.

      • Bart,

        Stop proving Voltaire right, cease praying to St. Jude the Apostle, and desist from encouraging Fred. It only increases his blood pressure.

        As for the rest it is nothing that we have not heard before, are very bored with and can’t be bothered responding to. Please learn a new song before posting next. Otherwise you simply come across as the increasingly bitter and alienated old man that people elsewhere are describing you as.

        Just a word to the wise
        CH

      • Chief

        Why would you address a word to the wise to me?

        And if you could identify which of your latest 84 words the wise one is? (Not that I doubt your ability to estimate this ratio.)

        As for your constraints, you want my replies to be novel, not boring, not bitter.. and wait, who said I sounded alienated and old? Could you link to that? I miss being alienated. It’s almost as if no one understands alienation anymore. Why, I remember in my day, when alienation was king! King! I say. Now.. where was I?

        Oh. Yeah.

        You want me to stop repeating my corrections, stop repeating your errors.

      • There is no right or wrong Bart – we are fighting culture wars as a result of divergent values. I value freedom, humanity and intellectual honesty. You value shooting off your mouth without engaging your brain.

      • That’s why I like you so much, Chief. ;)

  63. TomFP said “And while you are at it you’d better explain to Judith that, far from being Black Swan events, severe NH winters and Western Pacific floods were in fact all scenarios warned of by Climate “Science”.”

    Well, speaking only of Queensland, Chapter 11 of AR4 WG1, Regional Climate Projections, was very careful to make no specific projections for Australia and Queensland until 2080-2099 ((fig.11.17), by when only those under 30 now are likely to be alive to verify whether its actual prediction of NO FLOODS in the Western Pacific proved correct: ALL 21 of the models deployed to make that prediction actually forecast precipitation at LESS than the average in 1980-1999. The skill of the 21 models is such that they make no predictions before 2080.

    It is characteristic of Climate “Science” that it makes no predictions of anything within the lifetimes of those now aged over 30. That rules out inconvenient Black Swans giving those scientists a nip in their bums.

    Luckily that WG1 Chapter avoids making any specific predictions for frequency or intensity of Tropical Cyclones in the SW Pacific (p.915), so the recent cyclones in Queensland were not out of the ordinary.

    The kindest word I can find to describe Chapter 11 is that it is completely vacuous, aka an empty box.

    Now how about Australia’s greatest living Climate Scientist, Ross Garnaut (aka Roo because his 2008 Review recommended eliminating all ruminant livestock and relying only on kangaroos for meat, pp.540, 547-548), who is author of nothing less than the last word on THE Science of Climate Change, 2011, truly the definitive text in all its 66 text pages, containing ALL you need to know. For example

    1. “The temperature of air affects the amount of moisture it can hold and higher temperatures can lead to increased evaporation of water from the surface. The water-holding capacity of the atmosphere is expected to increase roughly exponentially with temperature rises” (Roo: 2011 p.26).
    For non-climate scientists, more evaporation means more precipitation, but not for Australia’s:

    2. “Two thirds of the 23 climate models used to inform Australian projections agree that rainfall will decrease in southern areas (for both the annual average and in winter), in southern and eastern areas in spring, and along the west coast in autumn (CSIRO and BoM 2007). In other regions and seasons less than two thirds of models agree on the direction of change, but in almost no region or season do more than two-thirds of models suggest an increase in rainfall” (CSIRO and BoM 2007). (Roo: 2011 p.28).
    Roo does admit that Australia’s rainfall is largely determined by ENSO, and the high rainfall across the whole country except for SW Australia in October 2010-March 2011 is associated with the current La Nina. But most of Roo’s 23 models imply there will never be another La Nina. So if there is that will indeed be a Popperian Black Swan.

    • TRCC,
      Do you have a link for those excerpts?
      I would like to use them, if you do not mind.
      Thanks,

    • This kind of makes the point that the IPCC said pretty much nothing about what will happen in the next decade or two. Whether the estimates of climate states in 2080 is meaningful or not is a different question.

      The citations from Garnaut don’t say anything about the time scales. Also the fact that 1/3 of the models differ in sign suggest that not too much trust should be put in them. BTW which 23 models is he referring to – GCMs, or regionally downscaled models for Australia?

  64. Hunter:

    The whole of AR4 WG1 is available via Google.

    Here is the link to Chapter 11:

    http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch11.html

    For the Roo Garnaut THE Science of Climate Change,
    go to
    http://www.garnautreview.org.au/update-2011/update-papers.html – Cached►

    Paper #5.

    Good luck! Roo and Madoff are birds of a feather.

  65. If one is looking to anticipate the cost of a Black Swan Event, then both cause and timing should be considered. It is a certainty that a massive volcanic eruption cools the planet; when it will happen again is essentially unpredictable. Such a happening should qualify as a BSE because its occurrence is unanticipated in the short-to-medium term.

  66. Spam filter’s being hyper again.

  67. The implications of black swans have been explored here close to the point of exhaustion, and so I hope I’ll be forgiven for introducing an OT comment relevant only in that it relates the color of birds to the existence of empirical data. Specifically, I’m intrigued by the principle that seeing a red tulip serves to confirm the proposition that all crows are black. I agree with that argument, but it appears to remain a subject of debate in discussions about the nature of evidence.

    • Brandon Shollenberger

      Not all crows are white, and seeing a red tulip in no way confirms anything about the color of crows.

      • Brandon Shollenberger

        I’ve seen a white crow.

        It didn’t look any too healthy.

      • Brandon Shollenberger

        There is an issue with what is meant by “black crow.” If you require a crow to be completely black, there are many non-black crows. In some species of crows, as much as 5% of the population can have white on them. They can have full-fledged albinism, which usually comes with health problems, or they can just have local issues (such as damage from trauma).

      • Brandon – In a practical sense, you are certainly right, but as a philosophical argument, the case is different.

        Scientific theories or hypotheses are almost never provable, but in scientific parlance, they are subject to “confirmation” in the sense of evidence that increases the probability they are correct. I used the word “confirm” in that specific sense, but not in the sense of “proof”.

        If we assume the existence of crows, two propositions are logically equivalent:

        1) All crows are black
        2) All non-black objects are non-crows

        It would be daunting to prove all crows are black by looking at all crows, but in theory it could be done (assuming the proposition is correct). However, seeing even one black crow is a piece of confirmatory evidence, because it reduces the number of opportunities for the proposition to be false.

        It would be even more daunting to prove all crows are black by looking at all non-black objects, but by the same reasoning, seeing even one non-black object that is not a crow would provide confirmatory evidence. The degree of confirmation would be immeasurably small, but the principle remains valid.

        I’ve heard that there are some crow species that are in fact not completely black, but the epistemic implications of the reasoning are still entertaining and perhaps informative.

      • Brandon Shollenberger

        The two propositions you list are not logically equivalent. They only become equivalent if one posits the existence of crows. Failing to do so leaves open the possibility of them being vacuous truths that contradict each other. You could theoretically observe every non-black object in the universe, but it wouldn’t prove crows were black since you wouldn’t have proof crows even exist.

        Now then, if you want to discuss observing a red tulip in addition to doing other things, it can be used as evidence. However, it is meaningless evidence as objects change constantly, and while you observe a single object, you’ve failed to observe hundreds of thousands more which have changed. Even worse, within a few weeks, that red tulip will most likely not even be a red tulip anymore.

        The only way observing a red tulip could speak to the color of a crow is if you could somehow halt time and examine the universe in stasis.

      • Brandon – Well yes, but as I stated, the two propositions are logically equivalent if crows exist. Crows exist, therefore the two propositions are logically equivalent. As to your other point, that the red tulip might change, I agree that if the red tulip mysteriously changes within a few weeks into a red crow, then my argument applies only to the current state of crow color. I have to admit that I failed to consider that possibility.

      • Also, although there is no practical value in trying to ascertain the color of crows by virtue of seeing a red tulip, the principle involves Bayesian reasoning that has substantial practical applicability in science – i.e., the notion that probabilities change on the basis of observed data, even when the changes are too small to be obvious. It’s the same reasoning that tells us that if we observe a coin flip to come out heads, we can conclude that its probability of coming out heads on the next flip has increased.

      • Brandon Shollenberger

        I wanted to cover everything, so I revisited a point you already mentioned (I was adding onto my comment, not just responding to yours). I probably should have worded that more clearly.

        However, your response to the main issue is completely lacking. Your flippant response misses the point in a nonsensical manner. It doesn’t matter what a tulip might change into. All that matters is the tulip changes. You offered the possibility of examining a single object to make a claim about the group to which the object belongs. If the object ceases to belong to that group, it can no longer be used to make any claims about the group.

        Now then, you could change “all non-black objects” to “all non-black objects which have existed or will exist at any time.” In that case, observing a red tulip would speak to the color of crows as long as existence is finite.

      • Brandon – Aren’t you pushing too hard? My original point remains valid – if crows exist, seeing a red tulip serves as confirmatory evidence for the proposition that all crows are black. Nothing said since that comment alters that conclusion. Is it confirmatory evidence that all crows will be black in the future? Well, actually it is, surprising as that may seem, although less strong than as confirmation for the existing state of crows. In order to reassure yourself that this is true, I think you should consider that a confirmatory piece of evidence does not “prove” a proposition, but rather eliminates one opportunity for it to be wrong. That is why seeing a red tulip now reduces the probability that some future crow will be non-black simply by virtue of reducing the probability that some current crow is non-black. The practical implications are irrelevant.

      • Also – sorry about the flippancy. Your point is well taken. What I wanted to emphasize about confirmation is that as long as an observation might have falsified a proposition, but instead is consistent with it, it can be considered confirmatory. In examining any piece of evidence, I believe that criterion should be kept in mind.

      • Brandon Shollenberger

        I have no problem with the concept. My problem is with your formulation of it. Failing to account for temporal issues renders your position invalid. Seeing a red tulip means nothing if you only consider currently existing objects due to the flux in the group being examined. You can salvage the concept by simply adding a phrase or sentence, but unless you do, your position is invalid. Missing premises are only allowed if those premises are immediately and inherently obvious. Yours weren’t.

        This may seem like nitpicking, but if your discussing philosophical issues, you should expect it. Specificity is demanded in fields like epistemology.

      • Brandon Shollenberger

        And I agree with the point you’re getting at. You just have to be careful not to draw false conclusions. It’s easy to think something is evidence when it is not, and it is even easier to double-count evidence. That’s why I have such an interest in specificity.

        Also, I would love an edit/preview feature. I can’t believe I typed “your” instead of “you’re.”

      • oneuniverse

        Fred: “That is why seeing a red tulip now reduces the probability that some future crow will be non-black simply by virtue of reducing the probability that some current crow is non-black.”

        Not really.
        Proposition A: All moons are made of green cheese.
        Proposition B: All moons are made of blue cheese.
        Proposition C: All moons are made of purple cheese.
        etc.

        Now begin to examine every speck of material in the universe other than that which constitutes the moons.
        According to you, as the evaluation progresses, the probability that the proposition A, and B, and C etc are true increases.

        Increasing the number of mutually exclusive propositions concerning the moons makes no difference. The probability of each of them being true increases.

        Since your scheme assumes some exhaustible probability space (otherwise the probability will not increase perceptibly), the probability of each proposition being true will approach one, or rather 1 minus (“no. of examinations involving moons” divided by”no. of possible examinations in universe”), which will be very close to 1.
        Since at most all but one of the mutually contradictory propositions will be false, the assignation of probability close to one of the others being true is shown to be erroneous.

      • Brandon Shollenberger

        oneuniverse, your complaint is incorrect. The propositions you listed are mutually exclusive. This means you cannot simply combine them in analyses. The objects you examine for Proposition A aren’t the same as the objects you examine for Proposition B, thus you’re comparing apples to oranges.

      • oneuniverse

        Hi Brandon, I don’t follow what you’re saying.

        The examinations are independent of what propositions are proposed. There’s no combination – Fred’s increase of probability occuirs in parallel, as it were, for all the propositions.

        Observing a red tulip will increase the probability of the existence of a non-black crow (according to Fred), and it will increase the probability of a talking crow etc.

      • oneuniverse

        Actually, the black / talking crow example wasn’t good, as they aren’t mutually exclusive. I chose mutually exclusive propositions to demonstrate the contradiction.

        Ok I see where you’re coming from (?)- but yes, the probabilities of the mutually exclusive propositions need to add to 1, if they are exhaustive. Since they’re all increasing towards 1, there’s a contradiction. Something is wrong if the probability of each proposition being true approaches 1, yet all but one (or all) are false.

        More simply, examining tulips provides no information about crows – it does not reduce the uncertainty concerning crows.

      • Brandon Shollenberger

        oneuniverse, once you accept they are mutually exclusive, it is a simple matter to see the probabilities do not all approach one. The subset of objects examined for all propositions is necessarily limited to those objects which fall in none of the categories (non-blue/purple/green). This means the probability cannot approach more than one divided by the number of propositions (plus one).

        If you literally examine all non-blue/purple/green objects as separate categories, by definition, you will have examined every object in existence. At this point, probabilities are no longer an issue as you’ll have examined the objects whose details you are trying to ascertain.

        I think you’re getting tripped up because of a lack of formalization. Try testing your belief by getting balls of different colors and setting aside one as a mystery ball.

      • oneuniverse

        oneuniverse, once you accept they are mutually exclusive, it is a simple matter to see the probabilities do not all approach one.

        That was my point – according to Fred’s notion, they do all approach something close to one, therefore there’s something wrong with the notion.

        The subset of objects examined for all propositions is necessarily limited to those objects which fall in none of the categories (non-blue/purple/green).

        I specified that the subset of examinations be limited to those objects which aren’t moons (or crows). This subset will be vastly larger than the subset of objects that are moons (or crows).

        Try testing your belief by getting balls of different colors and setting aside one as a mystery ball.

        I have N balls, each of some colour. The number of different colours and their frequency is unknown.
        I set aside one ball, say ball A, without knowing its colour.

        I then examine one of the remaining balls. Does that give me information about the possible colour of ball A. As I understand it, no it doesn’t. Similarly for the rest: I can examine all of them, and I still won’t be any wiser what colour ball A is.

      • Brandon Shollenberger

        oneuniverse, I hadn’t realized you didn’t include moons in your subset. That’s improper. The entire reason Fred Moolten’s approach works is because crows (in your case, moons) are included in the set of objects being examined.

        I apologize for not realizing your mistake sooner. You did state it in your initial comment, so I should have noticed it.

      • oneuniverse

        It’s not improper or a mistake, it’s using a particular sequence of examinations (all non-crows, or non-moons first) to highlight the absurdity. One can examine the crows or moons next, but it would be beside the point.

        According to Fred, one gets information about crows by examining tulips. Or, in your example of the coloured balls, one gets information about the colour of the mystery ball by examining the other balls. How? Please work through a concrete example for me?

      • oneuniverse – To test the proposition that all crows are black, one could look at four categories, but only two are informative. The informative ones are (A) all crows, and (B) all non-black objects. The non-informative categories are (C) all non-crows, and (D) all black objects. In your blue moon analogy, looking at all non-moons is equivalent to (C) – it is non-informative because it will overlook moons that happen not to be blue. For category (B) to apply, you would have to look at all non-blue objects. That would test whether a non-blue moon existed (i.e., a falsification). Note that looking at all non-blue objects would NOT test the proposition that all moons are green, because it would overlook blue moons, which would falsify the “all-green” proposition.

        The relevance of the red tulip resides in the fact that it is a non-black object that is not a crow. The fact that it is a tulip, if not referenced to its color, would have no confirmatory power (e.g., a black tulip would be non-informative).

      • oneuniverse

        Fred, category B (non-black objects) can be divided into two subcategories – B1, non-black crows, and B2, non-black non-crows.

        Examining objects in B1 gives you information about crows ( the objects in this sub-category, being crows, all belong to category A as well). Examining objects in B2 (eg. tulips) gives you no information about crows.

        So the only way to gain information about crows is to examine crows.

      • > As to your other point, that the red tulip might change, I agree that if the red tulip mysteriously changes within a few weeks into a red crow, then my argument applies only to the current state of crow color. I have to admit that I failed to consider that possibility.

        Goodman spotted a possibility that looks like that:

        http://en.wikipedia.org/wiki/Grue_and_bleen

      • oneuniverse

        So the only way to gain information about crows is to examine crows.

        Sorry, that should be “the only way to gain information about the colour of crows is to examine crows.”

      • Maybe it should be

        the best way to gain information about crows is to examine crows.

        Max

      • And the IPCC method:

        The IPCC way to gain (and disseminate) information about crows is

        a) start off by making a hypothesis concerning the color of crows and how this impacts their habits
        b) based on the hypothesis, create model scenarios to simulate the color and behavior of crows
        c) investigate fossil remains of pre-historic Archaeopteryx
        d) declare that the paleo-avain findings check closely with the model
        e) validate the model scenarios by comparing them with one another
        f) declare that the hypothesis has been confirmed by the model simulations
        g) project that (if unabated by exorbitantly costly measures involving major payments to dictatorships in the developing world) black crows will destroy all other species plus our planet by 2100
        h) ensure that “crow disaster scenario” is conveyed to public by a doomsday-hungry” media

      • Oneuniverse – I’ll give it one more try, and then decide that there is nothing further to be gained. Let’s go back to my original comment . Assuming that crows exist (and we know they do), the following two propositions are logically equivalent:

        1) All crows are black
        2) All non-black objects are non-crows.

        My first question is – do you disagree? If so, please cite a counter-example in which one of the propositions could be true while the other is false. I can more or less guarantee that to be impossible. So let’s proceed.

        If you want to determine whether all crows are black, you could examine all crows (proposition 1). Alternatively, you could examine all non-black objects (proposition 2). Do you not agree that either approach will answer the question? Specifically, if every non-black object is a non-crow, then all crows must be black objects (again, we presume crows to exist, which we know to be true).

        Now, let’s recall that “confirmation” refers to a finding that increases the probability of a proposition, without necessarily proving it (in science, hypotheses and theories are rarely provable but often confirmable). For simplicity, however, let’s assume that there are only four objects in the world rather than an enormous number, that two are crows and two are tulips, and that two are black and two are red. With that knowledge alone, we can estimate the probabilities (P) as follows:

        All crows are black (P = 1/6)
        All crows are red (P = 1/6)
        Some crows are black and some red (P = 2/3)

        Now, we stroll around and see a red tulip. This leaves three objects unobserved. Two are crows and one a tulip, and two are black and one is red. The red one could be either the tulip, one of the crows, or the other crow. The probability that all crows are black has now increased from 1/6 to 1/3, even though we haven’t seen a crow. We have”confirmed” the proposition that all crows are black.

        We stroll further and see another tulip. It’s red. At this point, we have observed all red objects, and we can now conclude that the proposition that all crows are black has been proved – again without seeing any crows. In other words, it is not correct to state that “the only way to gain information about the colour of crows is to examine crows”.

        I specified the numbers and probabilities as examples, but the principle doesn’t change even if we don’t have exact numbers. We wouldn’t be able to specify P values or to know how many non-black objects remain unobserved, but every time we saw a red (or otherwise non-black) tulip (or other form of non-crow), we will know that we have increased the P value for all crows being black. You could try this out with any specified distribution of objects and colors of your choice to calculate the probabilities before and after observing a non-black object, and I think you’ll find that the probabilities always change in favor of an all-black proposition for crows. On a practical scale, the change in P will be so small as to be insignificant, but as a logical exercise, it remains true.

        Note that this requires you to characterize every object your observe, which might or might not include crows. If you deliberately choose to ignore the color of a crow, should you happen to see one, or to refuse to look at any crows even when you encounter them, then you can’t draw conclusions about crows, because the confirmation only works if you are in a position to observe non-black crows but don’t see any.

        I’ll let you struggle with it for a while, but I don’t think there’s any way around the logic.

      • Fred,
        Your example is so obviously wrong that the point is really in telling, where the logic fails. It’s not difficult to realize that your logic requires implied assumptions that are not valid in reality.

        Your statement would be true, if we would have a method of picking randomly one red object of all red objects. Then every observation of non-crows would provide a very small additional evidence against the existence of red crows, but seeing a red tulip in real life has of course nothing to do with crows of any color. Seeing any red object has two effects that compensate exactly the influence on the existence of red crows: The observation reduces the expected share of crows among all red objects, but it also increases the expected number of all red objects leaving the likelihood of red crows unchanged.

      • oneuniverse

        Fred, you wrote: For simplicity, however, let’s assume that there are only four objects in the world rather than an enormous number, that two are crows and two are tulips, and that two are black and two are red.

        You’ve just added something that wasn’t present in what you originally said – in the world you’ve just described, we now know a priori that the number of red objects is equal to the number of black objects (or more generally, we know of a numerical relationship between them). Clearly this would help us improve probabilities through observation, but it wasn’t part of the scenario you originally described.

        Let’s remove that relationship, and make it simpler (I’ll expand to the 2-crow 2-tulip world at the end):

        There are only two objects – a tulip and a crow. Possible colours of each are red or black.

        Let’s say I set up the world by rolling a 4-sided dice.
        1 = red tulip, red crow
        2 = red tulip, black crow
        3 = black tulip, red crow
        4 = black tulip, black crow

        I can decompose my 4-sided dice-roll into two independent coin flips – coin flip one determines colour of the tulip, flip two the colour of the crow.

        You then come to the island and find the crow is red (or black) – you’ve effectively determined what my first coin flip was. What does this tell you about the result of my second coin flip determining the colour of the crow – precisely nothing, since the coin flips are independent. You’re still left with a 50% probability of the crow being red or black.

        We can expand this to 2-tulip 2-crow world by using a 16-sided dice to set it up, or 4 independent coin flips – 2 flips to determine the no. of black (and therefore red) crows, and 2 flips to determine the no. of black tulips. Again, determining the coin flips for the tulips by observing the number of red tulips won’t help you determine what the outcome of the other two coin flips for the crows.

      • oneuniverse

        I mistakenly wrote : You then come to the island and find the crow is red (or black)

        That should’ve been tulip, not crow – sorry.

      • Oneuniverse – see my reply to Pekka above, and the correct link below.

      • oneuniverse

        re: Hempel’s paradox

        Hempel’s paradox relies on the Equivalence Condition (EC): “If a proposition, X, provides evidence in favor of another proposition Y, then X also provides evidence in favor of any proposition which is logically equivalent to Y.” (quoting from the Wikipedia article).

        For the example 1-crow 1-tulip world above (and in general for the n-crow n-tulip worlds constructed in the same way), the EC does not hold:

        “all crows are black” (A) is logically equivalent to “everything not black is not a crow” (B).

        I showed that finding a red tulip (evidence for B) did not alter the probability of finding a black crow, (so was not evidence for A) since the coin flips deciding the two were independent. Therefore the Equivalence Condition does not hold for this logical situation, and is therefore not universal.

      • > I showed that finding a red tulip (evidence for B) did not alter the probability of finding a black crow, (so was not evidence for A) since the coin flips deciding the two were independent.

        And yet everything that is not black is not a crow, and every crow is black. Both conditions are still logically equivalent.

        The only thing that Fred added is that we know how many things there are in the world. This is not unimportant, but not because of the probability reasoning offered here.

      • oneuniverse

        willard: The only thing that Fred added is that we know how many things there are in the world.

        In his example above, he also added that we know that there are two red objects and two black objects.

        willard: And yet everything that is not black is not a crow, and every crow is black. Both conditions are still logically equivalent.

        What do you mean by “And yet..” (you go on to restate the propositions) ?

        I said that the propositions were logically equivalent but that the EC doesn’t hold for my example.

      • > In his example above, he also added that we know that there are two red objects and two black objects.

        That’s what I meant. This restriction actually reinforces counter-arguments based on probability. To “try to remove that relationship”, another restriction was added, according to which the confirmational basis of laws must be based on the probability to predict the next thing one encounters.

        This is simply not the same question anymore:

        > Hempel’s point was that the application of Nicod’s criterion means that, since even observations of non-ravens are confirmatory, the class of neutral instances in fact has no members.

        http://plato.stanford.edu/entries/hempel/#EquCon

        Even if I always have 50% chances to encounter either a tulip or a raven, if we’re to follow Nicod’s criterion, which is:

        > [I]n relation to conditional hypotheses, instances of their antecedents that are also instances of their consequents confirm them; instances of their antecedents that are not instances of their consequents disconfirm them; and non-instantiations of their antecedents are neutral, neither confirming nor disconfirming.

        the paradox obtains, at least according to Hempel, provided we **add** EC:

        > Given that logically equivalent hypotheses have the same empirical content, whatever confirms one member of a set of logically equivalent hypotheses must also confirm the others.

        The point is not to determine the results of future coin flips, but the color of crows. That is, one’s relevant laws of ontology are reinforced the more one encounters crows that are blacks or non-crows that are not blacks, provided we accept that logically equivalent hypotheses have the same empirical content.

        It should come as no surprise that logically equivalent hypotheses have the same empirical content is a very problematic idea.

      • oneuniverse

        willard: That’s what I meant.

        Ok – what you said was quite different.

        willard: To “try to remove that relationship”, another restriction was added, according to which the confirmational basis of laws must be based on the probability to predict the next thing one encounters.

        No – I removed the restriction that the no. of red objects = no. of black objects – that just changes the permissable distributions of colours amongst objects.

        The point is not to determine the results of future coin flips, but the color of crows.

        Coin flips were used only to set up the world, before the observations took place. There’s no need to determine future coin flips.

      • The sentence:

        > [T]he paradox obtains, at least according to Hempel, provided we **add** EC […]

        is the opposite of what it should read:

        > [T]he paradox obtains, at least according to Hempel, _unless_ we **add** EC […]

        The Stanford explanation is less clear than the original Hempel article. If I can’t find a copy on-line, I’ll try to quote the relevant parts of it if somebody shows interest.

      • oneuniverse,

        I believe you’re confusing what the dice thrower knows and what the tourists coming into your island knows.

        The fact that you, the dice thrower, knows that you are throwing a dice, and that each of your throws provides a 50% chance to tourists of discovering the color of a tulip or of a crow or whatnot, all this is irrelevant to what is discovering the tourist while probing your world.

        I also believe that you are building a very strange island. You are trying to construe a perfectly random world, without any regularities. This seems implausible, as scientists are usually trying to find the lawlike patterns of the world.

        Hempel’s point is to say that one can’t build a concept of degree of confirmation solely on the syntactical formulations of the laws. His paradox was there to show that, even if we want to portray science as a deductive logical structure, we need to **interpret** the empirical statements. His EC was there to show that the content of these statements should be somewhat independent from their formulations.

        His EC has its own problems, but to refute Nicod’s criterion of confirmation, Hempel’s paradox seems perfectly sound.

      • We should also remember this joke (this version is from Wikipedias “Mathematical joke”):

        An astronomer, a physicist and a mathematician are on a train in Scotland. The astronomer looks out of the window, sees a black sheep standing in a field, and remarks, “How odd. Scottish sheep are black.” “No, no, no!” says the physicist. “Only some Scottish sheep are black.” The mathematician rolls his eyes at his companions’ muddled thinking and says, “In Scotland, there is at least one sheep, at least one side of which appears to be black from here.”

      • oneuniverse

        I believe you’re confusing what the dice thrower knows and what the tourists coming into your island knows.

        Please see my post to Fred below disambiguating the two.

        I also believe that you are building a very strange island. You are trying to construe a perfectly random world, without any regularities. This seems implausible, as scientists are usually trying to find the lawlike patterns of the world.

        Fred was saying it worked for all distributions. Ok, let’s impose a pattern – say, every third crow is red. Examining tulips will give no information about this pattern – we need to examine the crows.

        In order to be able to gain information about the colours of crows by examining tulips, there needs to be a colour relationship between the categories of objects. If you don’t know that a priori, you’ll still need to examine the crows to see if such a relationship is supported by the observations. Or there might be no relationship.

      • Here is my latest comment in the thread initiated by Brandon as described above:

        To conclude: Hempel’s paradox is caused by looking at a badly defined problem and including in the analysis implied additional assumptions, which are different in each approach. Using different implied assumptions leads to different conclusions as is expected.

        This observation is significant as it tells that it’s impossible to use inductive logic, when we cannot perform well controlled experiments, where all needed assumptions are explicit and fully controlled. This ideal can be approached well enough in many laboratory experiments, but it cannot be reached in observational studies of real world.

        Brandon commented on Judith’s blog that he wrote his opening message here, because this issue is not significant for the climate discussion. That comment is not really true, as all empirical knowledge on climate belongs to that class of observations, where the difficulties implied by Hempel’s paradox are present, and present to a very significant degree. They are actually at the heart of the disagreement on AGW.

      • The difficulty of following the lengthy threads hit again. This message should be lower after a message by Brandon Shollenberger.

      • Sorry for the bad link. This should work:

        Hempel’s Paradox

      • Fred,
        Paradoxes are created by artificial rules that violate reality.

        Without artificial constraints that are not true for real world observations seeing red tulips has absolutely no connection to the existence of red crows.

        On the other hand it’s true that not seeing a red crow reduces their likelihood, if that happens under conditions, where crows might be seen, but what kind of tulips we see at the same time has no influence.

      • I should add that the arguments presented and attributed to Hempel and others in the Wikipedia article are just wrong.

      • Oneuniverse – I believe the fallacy in your example lies in the fact that you have a priori assumed as fact that tulips and crows are equally likely to be red. In that circumstance, it would of course be true that a random observation would be consistent with that scenario. However, the point of the problem is that we don’t know how color is distributed and are using observations to infer what the distribution might be. Under those conditions, seeing a red tulip is evidence (slight as it might be) that red objects are more likely to be tulips than crows.

        Let me illustrate with a simplistic example that utilizes only three scenarios, but the same principle applies if we consider all intervening gradations. In scenario 1, all red objects are tulips and all black objects crows. In scenario 2 (your scenario), red objects are as likely to be tulips as crows. In scenario 3, all red objects are crows and all black objects tulips. These are a balanced set of scenarios, which in the absence of observations, favor neither tulips nor crows. We now sample an object from the population of objects, and it is a red tulip. This observation eliminates scenario 3, leaving an imbalance (scenarios 1 and 2) favoring the blackness of crows. (Obviously, as mentioned many times earlier, it is a miniscule step rather than a proof). There are many other ways to frame the issue, but they lead to the same conclusions.

        I enjoy this problem for a few reasons. First, I like paradoxes and counterintuitive realities. Second, I find probability an intriguing concept, partly because it’s so elusive. Third, I’m impressed by the power of Bayesian analysis. Fourth, and somewhat egotistically, I’m pleased that when I first encountered the problem, I reasoned on my own to the same solution reached by Hempel and very many others over the course of 60-70 years – a history I was unaware of until I looked into the paradox further. That general solution now appears to be well established, although with quibbling at the margins regarding allowable assumptions about background information and other parameters.

        I believe you should seriously contemplate the long and extensive history of this paradox, as analyzed by experts in probability theory, Bayesian analysis, and epistemology. It is still conceivable that new chinks in the edifice will be discovered, but do you think that obvious flaws of the kind you are looking for would have escaped the notice of all these individuals for all these years, only to be discovered by you or others on the basis of fairly simple probability paradigms?

        Regardless of your answer, I think other readers intrigued by the paradox would be interested in visiting the link I cited above for some of the many details on this topic.

        Willard – Thanks for the link to the long article on Hempel.

      • oneuniverse

        Fred: I believe the fallacy in your example lies in the fact that you have a priori assumed as fact that tulips and crows are equally likely to be red.

        You wrote earlier : “You could try this out with any specified distribution of objects and colors of your choice to calculate the probabilities before and after observing a non-black object, and I think you’ll find that the probabilities always change in favor of an all-black proposition for crows. ”

        So there shouldn’t be an objection to my choice of colour distributions determined by coin flips.

        Also, my example was an adaptation of your example, which also assumed that tulips and crows are equally likely to be red (each object was 50% prob. red, otherwise you wouldn’t have got “All crows are black (P = 1/6), All crows are red (P = 1/6), Some crows are black and some red (P = 2/3)”).

        It’s not a fallacy. It’s a simple set of examples in which finding red tulips doesn’t increase the probability of all crows being black.

      • Oneuniverse – Sorry if my example confused you, but it was different from yours. In my example, the probabilities for distribution of color were not an established precondition, but were simply estimated before a sample was taken, and therefore subject to change on the basis of the observed sampling – they weren’t fixed, and did in fact result in an increased probability of black crows after sampling. In any case, please feel free to disregard the example I cited, which was only roughly analogous to the problem at hand, and seriously review everything else in my comments and more importantly in the linked article. You may well find that the “black raven paradox” comes up again from time to time and so you would probably want to be prepared with as accurate a perspective as possible.

      • Perhaps the problem resides on the use of “probability”.

        Here is what Hempel says:

        > [W]e have adopted the more neutral term ‘degree of confirmation’ instead of ‘probability’ because the latter is used in science in a definite technical sense involving reference to the relative frequency of the occurence of a given event in a sequence, and it is at least an open question whether the degree of confirmation of a hypothesis can generally be defined as a probability in this statistical sense.

        Also note that in the introduction, Hempel says that he is trying to “provide the elements” of “a theory providing general criteria of confirmation and disconfirmation”.

        The take-home of his paradox would be that:

        > [A]n adequate definition of confirmation will have to do justice to the way in which empirical hypotheses function in theorical scientific contexts such as explanations and predictions […]

        There are lots of things that can be added, but this should be enough to shed some light on the different between Fred and omnologos.

      • Willard – Whether “probability” and “degree of confirmation” are synonymous can be argued – they probably aren’t – but for this particular paradox, I believe they operate in the same direction. Quantitatively, we might not be able to give a defined value for the probability increase of black crows (or ravens), but I think we can still state that they are more probable after seeing a red tulip, and it is hard to visualize an increased degree of confirmation associated with a reduced probability of that particular conclusion. Unless a counter-example is demonstrable, I probably won’t try to expand on this, because these exchanges are already very long. It’s certainly an interesting subject, though. I appreciate your insights and the Hempel reference, which addresses the concept of confirmation in a much more general way than simply the application to this particular problem.

      • oneuniverse

        Fred, you wrote :
        “You could try this out with any specified distribution of objects and colors of your choice to calculate the probabilities before and after observing a non-black object, and I think you’ll find that the probabilities always change in favor of an all-black proposition for crows. ”

        My choice of distribution of objects was equal numbers of crows and tulips, and my choice of distribution of colors was picked by flipping a coin for each object.
        This is within your scheme, and according to you, finding a red tulip should increase the probability of all crows being black, but in my example it doesn’t.

        Fred: I believe the fallacy in your example lies in the fact that you have a priori assumed as fact that tulips and crows are equally likely to be red.

        Perhaps there’s some confusion about which probabilities we’re talking about. There’s the 0.5 probability involved with the coin flips, and the probability from the observer’s viewpoint when sampling.

        The coin flips determine the distribution of colours amongst objects, but the generated distribution will not mean that for the observing sampler tulips and crows are equally likely to be red. For all possible distributions generated in this way, finding a red tulip doesn’t change the probability for the observer of finding a red crow.

      • Here is the link to the original article:

        http://www.jstor.org/stable/2250886

        Another source:

        http://www.hss.caltech.edu/~franz/Confirmation%20and%20Induction/Carl%20Gustav%20Hempel.htm

        Let me know if you need any help reading it. You can find my email in my tumblog.

      • Willard – Thanks for the links. Oneuniverse – The coin toss analog is invalid. You’re making the same mistake as before in assuming that what’s left after you remove a sample has the same probability as before except for the subtraction of the sampled item. It doesn’t. I’ve given you the best suggestions I can for review. You can use or ignore them as you see fit, but please take seriously the improbability that you’ve discovered an obvious flaw in the paradox that everyone else has overlooked for 60 years. This set of exchanges is already too long, but if you want to email me, you can find my email via my website or the denizens page.

      • I find it unbelievable that this simple and totally artificial example has become a widely discussed paradox.

        It’s to me absolutely clear that there is no paradox of this nature in any consistent framework and that the way the inconsistency is built to this paradox is so plainly visible that the whole issue should be clear to everybody.

        The apparent paradox is created by the fact that we have two similar but different problems that are not kept separate, but mixed to each other.

        The first problem is the real world problem, where red tulips and red crows are separate independent issues. We all have the right intuition on this problem: Because there is no physical connection between red crows and red tulips, no observation on red tulips has any slightest influence on the number or probabilities of red crows. This solution is absolutely correct and no amount of wrings by Hempel or others can cause any doubt on its validity.

        How can we then be led to those erroneous reasonings that Fred and Hempel have presented. That is done by introducing to the analysis assumptions that are false. They introduce a concept of all red objects and start to make conclusions based on that. They do not notice that this new concept is an union of separate sets which include red tulips and red crows. They do not notice that we have direct information and justified priors in Bayesian approach for these subsets, not for the union. Everything we know about the union is obtained combining, what we know about the subsets. Thus making an observation on the subset of red tulips changes both our knowledge on the subset of red tulips and the union, but leaves the knowledge on the subset of red crows unchanged.

        It’s just unbelievable that this obvious resolution of the false “paradox” is not known and accepted by everyone.

      • Looking at Hempel’s article, it becomes clear that the “paradox” is really an indication of difficulties in inductive logic or “The Logic of Confirmation”.

        We know that looking at tulips doesn’t tell about crows. There is no problem there, but there is a problem in defining rules for logic of confirmation.

        Concerning the claim “all crows are black”, we may even ask, whether seeing one black crow provides confirmation for the hypothesis. It does, if we have picked a randomized sample from the collection of all crows, but under some other settings, the observation might not be of any confirmatory value. Similarly picking a truly random representative from the set of all non-black objects would provide confirmation for the hypothesis (or disprove it), but without the procedure of random sampling from the full set, we may have no confirmatory value at all.

        The case of observing a red tulip in absence of additional constraints falls to this case of exactly zero confirmatory value.

      • oneuniverse

        Fred: The coin toss analog is invalid.

        Why is it invalid? You said this would work for any colour distribution I chose, so I’m choosing by flipping coins. Fairly standard stuff.

        Fred: You’re making the same mistake as before in assuming that what’s left after you remove a sample has the same probability as before except for the subtraction of the sampled item. It doesn’t.

        It’s not an assumption. You can verify my example by doing repeated runs of the experiment.

      • Brandon Shollenberger

        I’m a bit disturbed to see how long this off-topic discussion has lasted. I don’t see any indication progress has been made, so I suspect me joining in again would just add more clutter.

        If there is a more appropriate place to discuss this topic, I have a semi-formalized solution written which I’d be happy to share. I’d post it somewhere and provide a link, but I don’t have anywhere to host it.

      • brandon i’ve lost track of what is going on in this particular discussion, if you have a topic to recommend, send me an email

      • Brandon Shollenberger

        I don’t think the topic being discussed would make for a good blog post here. It’s a mildly interesting topic, but it has nothing to do with the climate, global warming, or anything related to this blog.

        That said, Pekka Pirilä has a page on his blog where he’s invited discussion on this matter, and I’ve posted an explanation there. I think my comment could help resolve the disagreements people have been having here.

        I’d invite everyone who has participated (or just is interested) in the discussion to go read it and post any concerns they might have.

      • The argument is valid. The methodology is poor.

  68. Scare mongering every 60-years!

    Melting Polar Ice Caps to Raise the Level of Seas and Flood the Continents
    http://nyti.ms/k9vsPT

    We still speak of “The Ice Age” as if it belonged to the remote geological past. Geologists have reached the conclusion that there were several Ice ages. What is more, the last Ice Age, known as the Quaternary, is only about half over, despite our blistering Summers. “Eternal ice” or “eternal snow” are figments of the poetic imagination. Very slowly the great ice sheets in the Arctic and Antarctic regions are melting and pouring their torrents into the oceans. The earth must inevitably change its aspect and its climate.

    How the change is slowly taking place and what the result will be has been considered by such able geologists as Professor Sir Edgeworth David of the University of Sydney, Australia, Professor Wilhelm Meinardus of Gottingen and a score of others. The latest is Dr. William J. Humphreys of the United States Weather Bureau, who recently addressed the American Meteorological Society on the subject, summarizing old views and modifying them in the light of the information gathered in the Antarctic regions by the Byrd expedition and in Greenland by the ill-fated Professor Alfred Wegner and his companions.
    ….
    The earth is steadily growing warmer. As all the ice at the two poles melts a stupendous volume of water will be released. Professor David Conservatively estimates that the sea level will rise fifty feet [15m!]. Professor Meinardus doubles that estimate. Dr. Humphreys, with the studies of Byrd and Wegner before him, believes the rise will be 151 feet [46m!]. Such floods are nothing new, as we see by the marine fossils found on the tops of the Rockies, Andes and other mountain ranges.

    • “Such floods are nothing new, as we see by the marine fossils found on the tops of the Rockies, Andes and other mountain ranges.”

      Measuring sea level in this manner requires an assumption that we knew the height of these locations accurately at the time of the flooding.

  69. When the debate is reduced to red tulips and black crows, it’s apparent that boredom has once again set in. Which may indicate nothing more than that Black Swans are recognizable only in retrospect and that anticipation of the event is a waste of time and energy.

    OTOH – a large part of spacecraft operations (as with military operations) is contingency planning. Not that one expects the situations being planned for to actually happen in the exact manner that’s planned for, but the planning process exposes the operators to the possible range of options available for corrective action when that Black Swan raises it’s ugly head.

    As General Eisenhower said during WWII – “Plans are useless, planning is essential.” And he should know, considering the number of Black Swans he had to deal with during that time.

    • Jim

      The number of times I’ve heard the exclamation, “There’s a plan for this!?”

      ;)

      • Yup – but did they follow the plan? And if so, how long before the plan failed?

        As someone once said – “No plan survives first contact with the enemy.”

        Witness NOLA and Katrina – they had a plan, too.

      • Jim

        Most plans don’t survive first contact with friendlies, either. :D

  70. Professor Curry,

    What is your take on this?
    http://www.warwickhughes.com/icecore/
    Note his later paper “CO2: The Greatest Scientific Scandal of Or Time”
    Was this “quibble” ignored by IPCC?

  71. Steven Mosher and Bart R

    IPCC:

    For the next two decades, a warming of about 0.2°C per decade is projected for a range of SRES emission scenarios. Even if the concentrations of all greenhouse gases and aerosols had been kept constant at year 2000 levels, a further warming of about 0.1°C per decade would be expected

    http://bit.ly/caEC9b

    Here is the observed global warming rate compared to IPCC projections
    http://bit.ly/iEU7Hn

    BLUE=> IPCC projections of a warming of 0.2 deg C per decade.
    VIOLET=> IPCC projections of 0.1 deg C per decade for CO2 held at 2000 level
    GREEN=> Observed warming of only 0.05 deg C per decade.

    It is futile to hide behind uncertainty instead of questioning AGW

    • Correction:
      Here is the observed global warming rate compared to IPCC projections
      http://bit.ly/lQbg3x

      BLUE=> IPCC projections of a warming of 0.2 deg C per decade.
      VIOLET=> IPCC projections of 0.1 deg C per decade for CO2 held at 2000 level
      GREEN=> Observed warming of only 0.05 deg C per decade.

    • Girma

      Technically, I’m hiding behind not caring.

      +AGW. -AGW. 0 AGW. Doesn’t especially matter to me, though it seems when all the data and analyses are laid out, +AGW is likeliest.

      It just doesn’t matter that much.

      Incompetent analysis of illegitimate claims or illegitimate analysis of incompetent claims, really, how are you imagining I should care to play such a game? Even competent analyses of legitimate but relatively unimportant claims don’t turn my crank much.

      Claims of forecasts based on the mean of models, when the models are divergent and sensitive to initial conditions, and exclude specifically known real significant factors tells me the person making the claim hadn’t thought very clearly before shooting themself in the foot. (But at least they were willing to consider mechanisms in their hypotheses, a pretty decent requirement of a scientific forecasting.)

      So using the best tools available, undauntingly honest hard work, and brilliant analyses to dissect such claims, as it appears Steven Mosher endorses Lucia for doing.. seems a waste of brains that could be doing something better, with all due respect.

      Then again, some people waste their time playing poker or chess or tic tac toe, so who am I to criticize? At the very least, Lucia is documenting decent technique online, far more valuable than the actual conclusions, IMHO.

      I won’t waste my time contrasting Lucia’s excellent (though I think wasted) example against other sharply contrasting (though definitely still wasted) examples at hand.

      • Bart R

        Then again, some people waste their time playing poker or chess or tic tac toe, so who am I to criticize?

        I have 24 hours a day. Eight hours for work, eight hours for leisure and eight hours for sleep. If I spend some of leisure time playing my beloved chess (the best game in the world) am I “wasting” my time? What is life other than doing what we love?

      • Girma

        I think it wonderful, and a very good thing, that you play and enjoy chess, or backgammon, or snakes and ladders or whatnot; if you love your game of choice, so much the better.

        Playing games with one’s claims in science, however, is an entirely different matter, to those of us who love science or truth.

        Perhaps a sports metaphor will explain a small fraction of my dismay.

        “Moving the goalposts,” ie changing the standards after the game has begun, is generally considered unsportsmanly.

        When the IPCC first made its (ridiculuous) claims of a 0.2C/decade temperature rise ‘forecast’, it was in FAR in 1990; while it repeated the claim in later reports, refined them, explained them, it fixed that (rather silly) claim in 1990.

        Move the goalposts to start your comparison in 2001 or 2002? Why? That’s like calling the score of the game based only on one of the middle periods of play.

        Further, the IPCC in both 1990 and 2001 generally disparaged claims of trends based on less than 30 year periods.

        The game they entered, the score was built on three decades of data.

        Even within the 0.2/decade claim as made, it was ‘for a century’, which we may wish to consider when considering the span of the claim.

        Did the IPCC intend to suggest there was so little variability in weather that no decades for a century would lie below their multi-model mean?

        Given that using that particular multi-model mean was absurdly ill-advised, they might have meant something so ludicrous, but why ought we share in ludicrous reasoning?

        Use the same decadal trend line logic to analyze temperature data back a century, and we find more than one decade with an actual negative slope where the 30-year trend is positive and it is generally agreed the overall temperature trend is rising. As the convention has been established, moving the goalposts in this way is a clear foul, a transparent attempt to stack the deck post hoc.

        So while it’s fun, and plausible, to say the IPCC was off by a factor of four in their +/-AGW prediction, it’s far more accurately sporting to say it appears that in a decade we will know if the trend of missing their +/-AGW prediction by 17% is borne out.

        Likewise, we have the Pro From Dover problem.

        The IPCC’s public claim appears to have been a simplification.

        Any competent analyst could dismantle it from the day it was made, in a strict reading of what had been said. I mean, I can dismantle the claim, based on no more information than was available when the multi-model mean was first proposed, and I’m not claiming to be all that competent.

        But the claim clearly wasn’t for the foxy Professional ranks; it was for the hedgehog spectators on the sidelines. In answer to what is the best forecast that can be made, after repeatedly being told that no forecast is possible, it looks like some tomfool within the IPCC fell into the trap of thinking that saying something is better than saying nothing.

        So while it’s right to disparage the tomfoolery, it’s an abuse of perfectly good statistical tools and analyses to apply them to this prediction; it’s a ridiculous prediction from the outset, why smear good statistical methods with the bad premise that they apply here?

        Climate is jello, not a billiard ball. Chaos dominates, not caroms.

        Misleading people by claiming you have a billiard-stick that shoots jello will ultimately do you no more good in the long run than when the IPCC made that same mistake.

        There is some, very limited, literature suggesting some promising methods may be used to handle spatiotemporal chaos within linear predictive systems, but it is early days yet on this, and no one, to my knowledge, has yet applied and proven such techniques in the +/- AGW case in a fully satisfactory way.

      • Move the goalposts to start your comparison in 2001 or 2002? Why? That’s like calling the score of the game based only on one of the middle periods of play.

        …it is estimated that they will increase by 0.2C every decade over the coming century
        http://ind.pn/i2ZHaw

        Observation compared to projections:
        http://bit.ly/lQbg3x

        BLUE=> IPCC projections of a warming of 0.2 deg C per decade.
        VIOLET=> IPCC projections of 0.1 deg C per decade for CO2 held at 2000 level
        GREEN=> Observed warming of only 0.05 deg C per decade.

      • “The trend in the ENSO-related component for 1999–2008 is +0.08±0.07°C decade–1, fully accounting for the overall observed trend. The trend after removing
        ENSO (the “ENSO-adjusted” trend) is 0.00°±0.05°C decade–1, implying much greater disagreement
        with anticipated global temperature
        rise.P/1.”

        “Near-zero and even negative trends are common
        for intervals of a decade or less in the simulations, due to the model’s internal climate variability. The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate.hV HT”

        What Girma is arguing isn’t that far off from what is argued by the scientists at the NOAA. I’m not sure why his technique draws such ire but if you go with what the experts are saying it really doesn’t change the goal posts by very much.

        http://www1.ncdc.noaa.gov/pub/data/cmb/bams-sotc/climate-assessment-2008-lo-rez.pdf

      • steven

        Girma (and others) appear to me to be pointing at very short (under 15 year) trends and proclaiming the existence of even one such trend in the actual data (although it has been published that such trends also appear in the models) as disproof of the hypothesis.

        This is in statistical analysis an illegitimate conclusion.

        If one has a confidence level of 95% that one will not see a fifteen year negative trend, and one finds one such independent trend in twenty, well guess what?

        One has exactly matched the prediction.

        There haven’t yet been 15 years since 2001, but even hindcasting to include the running 10-15 year trend lines between 1990 and today, one doesn’t develop sufficient exceptions to the rising trend to declare a falling or level trend overall by that standard. We’d been through that discussion months ago here.

        We do have almost enough information now to possibly reject that this decade will not rise between 0.2C and 0.5C (the model prediction) based on 30-year trends. (That is, it would have to get very hot very fast for the next 15+ years to accept the past decade as part of a 0.2C-0.5C rising trend. Far hotter, far faster, than we could reasonably expect.)

        This again would get us to only one independent decade that would not fall within the range predicted by the hypothesis as represented by the very silly IPCC statement of 0.2C-0.5C per decade for a century.

        Which brings me again to the multi-model fallacy.

        It is possible to improve predictions of trends by averaging them, sometimes*.

        Such averaging may improve the correlation of predictions to outcomes, but will almost always worsen correlation of derivatives. The central tendency is not kind to tangents. That’s one of the products of taking the mean: curve smoothing.

        If someone had wanted to produce a prediction based on multiple models that matched short-term trend lines, one would have had to somehow aggregate the short term trend lines of the models in a meaningful way, which is pretty much the opposite of what one gets by taking the mean of the models.

        Further, if we want our prediction to match both temperature and trend, that ‘meaningful aggregation’ technique would need the prescience to start and end its extreme trend lines at the same time as the real data. Which is simply magical thinking run amok. It ain’t gonna happen.

        You put powerful statistical tools into the hands of someone who doesn’t think through the implications of applying the tool, you will sometimes get such persuasive-seeming but nonsensical outcomes. It’s like handing a car to an irresponsible teenager.

        *Is the system sensitive to initial conditions? Do the predictions tend to converge? Does ergodicity emerge in the models? Do the models account for all significant factors? Then why take their mean?

      • Bart, I agree with your multi model fallacy. Each model should be evaluated seperately on the merits of the output. Girma isn’t the one that made the parameters of the argument. He is arguing against the argument presented.

        I agree that 10 years is too short a time to falsify the models. The models clearly show this. It is also clear that a period of 30 years is not required. It is also quite clear that the start date of any trend where you can find 15 years of no warming meets the qualifications set forth.

        I’m not sure what discussion you had months ago. In this discussion it seems clear to me that Girma is using too short a period of time. You are using too long a period of time. Girma is appropriate in starting the trend line where he chooses.

        As far as expecting as expecting to see a 15 year trend with no warming out every 20, I think you misread the paper. In fact the 10 year trends of no warming came out to about 5% of the possible trends and a 15 year trend of no warming is indicative of model falsification. Pielke Jr figured out how often the 10 year trends occurred here: http://rogerpielkejr.blogspot.com/2009/07/noaa-explains-global-temperature.html

      • steven

        Thanks for the links.

        I’m still expressing poorly what I have to say.

        Prefacing all with, applying statistical analyses to these curves is something I’m patently uninterested in pursuing for the sake of establishing anything about +/-AGW (not just because it’s almost impossible to do so with these methods), I’m glad to use them as the example at hand of what I find aesthetically and rationally wrong with the approaches presented.

        Models are better at producing ‘typical’ projections than ‘correct’ predictions.

        That is, a model will have features typical of the system simulated, depending usually on what typical features were considered in building the model, rather than the ability to string together those typical features in a predictive way.

        We expect models of complex systems under spatiotemporal or temporal chaos to have this flaw.

        Even if we well-understand mechanisms and integrate them into our hypotheses, sensitivity to initial conditions will affect outcomes over time as any one extreme event will derail the model for a span of time, unless there is some strongly convergent and immediate ergodic mechanism constraining this divergence.

        And here we get to the hedgehog-fox part of the show.

        I’m not a fox.

        I’m not a hedgehog.

        I believe the entire meme to be a faulty framework that leads nowhere.

        However, if there is anything to the hedgehog allusion, in chaotic systems a hedgehog can rationalize away all divergence from their prediction by simply saying “Aha! Extreme event acted differently here. I’m still right.”

        Which is just an invitation to me to beat my head on my keyboard. Linear analyses are generally inappropriate to chaotic systems when applied in this way; there may be appropriate applications, but they are very constrained, or not yet well-developed, or I just don’t know them well-enough to use yet because I left that study decades ago.

        So it’s not surprising to see the multi-model mean utterly stumble on a decadal scale coming out of the gates, and it’s not accurate to say that this stumble disproves the +AGW hypothesis.

        Per the NOAA:

        The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate.

        Which is where I got my 1 in 20 (the 95% confidence level) tolerance for under 15-year no-warming trends; this was of course a highly handwaved contrivance suitable to the context but not very complete. Roger Pielke Jr.’s analysis could get us to a better understanding, once he corrects some blatant errors in his assumptions (though he could have gotten fancier and gone farther, I’m sure, with some imagination and the time to waste on a moot point) for general application.

        Here’s another way of looking at it:

        http://www.woodfortrees.org/plot/hadcrut3vgl/to:1930/plot/hadcrut3vgl/from:1930/plot/hadcrut3vgl/to:1930/trend/plot/hadcrut3vgl/from:1930/trend/plot/hadcrut3vgl/from:1940/to:1950/trend/plot/hadcrut3vgl/from:1957/to:1967/trend

        (Forgive the messy graph.)

        Isolate your graph not into hypothetical trends (as, I hope, we can discredit that approach in the next paragraph), but into regimes where we claim the Range of temperatures belong to different climates. In this case, I’ve simply split the observations into roughly even halves.

        While the trend lines may be quite persuasive, and a useful shorthand for the characteristic direction of the temperature, we see that over and over (up to almost 1/3rd of all running decades) in the clearly rising trended climate post 1930 a negative or non-(unambiguously*)-rising trend (*within error bars). Granted, the trend of this period is a mere 1/3rd of the IPCC prediction of 0.2C/decade, but what this tells us is the presence of decadal trends might tell us almost nothing about the climate regime we are observing.

        That 17 of 61, or whatever the disjoint ratio of negative decades may be in the models, remain for the 0.2C-0.5C projections, compared to about 1/3 for the 0.07C climate regime from 1930 to today, might suggest that we could build a case for decreasing frequency of negative decadal trends in a more rapidly warming climate, but someone would have to sit down and prove that, as I’m a skeptic.

        More useful, suggested by the climate regime model, is the question, “Do the temperatures observed belong to the climate regime before 1930, or a climate regime warmer than the one from 1930 to 2001?”

        It wouldn’t take much work, I think, to conclude the temperatures are more like a regime warmer than the later of the two regimes than like the earlier (status quo) regime.

        Amplitude, not trend, we do know how to do the statistics for, in this case, to a fair degree of confidence.

      • Bart R

        …will ultimately do you no more good in the long run than when the IPCC made that same mistake.

        We skeptics will back off as soon as they admit their projection and policy prescription is not yet justified based on the data so far. It is extremely callous to scare billions of the world with climate catastrophe at this stage when the effect of human emission CO2 in the global mean temperature is negligible.

        Why negligible?

        Because the global mean temperature pattern before and after mid-20th century, before and after huge increase in human emission of CO2, are identical as shown in the following graph.

        http://bit.ly/iUqG8I

      • Girma

        Look again.

        The amplitude of temperature after 2000 never goes below 0.2C on your graph. (Of course, your graph is somewhat compressed, so we know curve smoothing aside, the global temperature did on occasion in this Domain drop below that mark, however we can consider these outliers for the moment.)

        The amplitude of temperature in the first half of the graph otherwise (1850-2000, say 1925) — let us call it ‘status quo’ for now — never rises above -02.C. (Likewise, compression is to blame for us not seeing the ‘outliers’ that do rise above this Range on the ‘status quo’ Domain pre 1925.)

        Can we accept the Range of temperatures post 2000 as being the same range pre-1925? No. We can fairly resoundling reject the ‘status quo’ or Null Hypothesis.

        Can we accept the Range of temperatures post 2000 as being higher than the Range for the transition period, the climate regime 1925-2000?

        We can’t yet, I believe, reject the hypothesis that these temperatures post-2000 are in a different Range from 1925-2000; but, if it is a different Range, it’s a higher one, that much we can tell.

        So where does this leave us?

        We expect 17 in 61 (or whatever) 10 year cooling trends if the hypothesis that the +AGW models reflect the actual climate, and we’ve encountered.. one of those, two?

        Fifteen or sixteen (or whatever) more to go to prove the climate is no hotter than the +AGW models predict, if we wait 689 more years and a bit and all other things remain equal.

        Well done us.

        Your arguments are not skeptical enough of your own assumptions and methods.

        Your conclusions are invalid due this lack of skepticism on your part.

        And skeptics ought never ‘back off’.

        Of course not.

        What would happen if apathy and complacency set in?

        We’d become beach bums living on the proceeds of lawsuits against fast food companies in trailer parks complaining the government didn’t do enough to protect us from big oil and bad weather, and was letting gas prices get too high. (No offense to any demographic I’ve inadvertently left out of this broadsided caricature.)

        However, your arguments by assertion carry little weight with me.

        You keep extolling virtues of vital fluids and the number 60 like some character from Dr. Strangelove, without substantiation other than pretty drawings that — if they mean anything at all — don’t mean what you claim.

        There may be some virtue to analyses of trends; however we’re still largely in the dark about climate trends and what they mean and what mechanisms may explain them.

        The 30-year minimum span for analyzing climate regimes may have no other basis than observed success, but it would be hubris to abandon a generally successful parameter for lesser and too-unexamined constraints.

      • Meh.

        Replace paragraph above exchanging ‘the same’ for first instance of ‘a different’:

        We can’t yet, I believe, reject the hypothesis that these temperatures post-2000 are in the same Range from 1925-2000; but, if it is a different Range, it’s a higher one, that much we can tell.

      • Bart R

        …will ultimately do you no more good in the long run than when the IPCC made that same mistake.

        We skeptics will back off as soon as they admit their projection and policy prescription is not yet justified based on the data so far. It is extremely callous to scare billions of the world with climate catastrophe at this stage when the effect of human emission CO2 in the global mean temperature is negligible.

        Why negligible?

        Because the global mean temperature pattern before and after mid-20th century, before and after increase in human emission of CO2, are identical as shown in the following graph.

        http://bit.ly/iUqG8I

      • Bart R and Girma

        Pardon me for intruding but something Bart wrote caught my eye:

        +AGW. -AGW. 0 AGW. Doesn’t especially matter to me, though it seems when all the data and analyses are laid out, +AGW is likeliest.M

        As Girma’s charts have shown, +GW is demonstrated by the GMTA record (assuming this record is valid). The underlying rate of warming has been about 0.6C per century since the record started in the latter 19th century. Superimposed on this trend line is a cyclical curve resembling a sine curve, with an amplitude of somewhere around +/- 0.2C and a total warming/cooling cycle time of around 60 years.

        This curve is statistically speaking a “random walk”, with no robust statistical correlation with atmospheric CO2, which has seen no cycles but has increased at a fairly constant CAGR of around 0.4% per year since measurements started at Mauna Loa in 1958 and at an estimated somewhat slower rate before this, based on ice core data. There appears to be a better correlation with ocean currents, such as ENSO, PDO, etc, and a fairly good long-term correlation with solar activity, but no mechanism has as yet been conclusively demonstrated AFAIK.

        So whether or not there has been perceptible +”A”GW or not is not firmly established (although IPCC would claim it is, based on theoretical deliberations and model simulations plus some questionable interpretations of distant paleo-climate data).

        But let’s see if we can agree with our hostess here (JC) when she says (as she did in testimony before a committee of U.S. Congress last fall):

        Anthropogenic climate change [also known as AGW] is a theory whose basic mechanism is well understood, but whose magnitude is highly uncertain.

        That would appear to me to be a pretty good summary of where the whole premise of +AGW stands today.

        Max

      • Max

        With respect, some quibbles:

        1. Girma’s charts aren’t really rock-solid proof; there is better evidence elsewhere, more clearly demonstrated and with more sound methodology.

        2. Girma has actually rejected claims of others who produced slightly different sine-based curves with different than 60-year periods (despite substantially better correlation coefficient of 0.92 vs. his 0.88), saying only let the data be the judge.

        Neither curve is backed by proposed mechanism, and both curves fall apart under even the most preliminary analytic tricks.

        Split the data into NH and SH, and see how quickly all sine curves degenerate or disappear and their correlations drop. This shouldn’t happen for a truly global phenomenon and indicates Simpson’s Paradox is responsible for a phantom observation.

        Alternately, compare the tangents of the sine curves proposed to the tangents of the smoothed real data, and notice these derivatives disagree all over the place; this ought not happen if the underlying data belongs to the hypothetical cycloids.

        Alternately, move the length of the period around from cycle to cycle and see if the correlation improves; it does, which ought not happen if the underlying data is truly hypothetical.

        Alternately, curves are proposed with both significant correlation and the necessary element of a plausible mechanism, like the stadium wave hypothesis recently mentioned in another thread. Comparing a viable stadium wave hypothesis with an unfounded cyclic one, we’re obliged to dump the cyclic hypothesis.

        Further, look at what the hypothesis is proposed to do: it’s claimed by the mere existence of this hypothesis that all +AGW hypotheses are discredited.

        Which just isn’t how it works.

        Even if there were only one hypothesis stated, we don’t take lack of statement of other hypotheses as proof of correctness; if a second hypothesis is stated, we don’t take existence of a second as disproof of the former.

        There may be infinite stated or unstated equally valid hypotheses. We can only accept or reject hypotheses on their own merit, though we may be persuaded to believe in the one of multiple alternatives that best satisfies our criteria at any one time for the purposes of policy.

        And.. Girma’s 60-year fixed period cuckoo’s egg simply doesn’t meet that criteria either.

        Your interpretation of Girma’s offering, while it is more rational in avoiding the Platonic Numerological implications of the number 60, and suggesting some appreciation of a non-periodic property to the data’s warming-and-cooling biases, likewise doesn’t quite get us there.

        For example, you reject robust statistical correlation with CO2, but neglect the mechanisms suggested to mediate CO2 with +AGW: it’s not a direct mechanism, is it?

        The Sun is involved, other CO2E gases, particulates, weather phenomena, ocean gyres, plants, plankton, albedo, sea ice extent, salt concentration of sea water..

        They all mediate to some extent.

        At best it’s something like a fourth or fifth order relationship or if more direct, also a function of something like six multiple variables, and that function iterative and complex, so spatiotemporal chaos involved in it.

        So it would surprise the heck out of me to see a statistically robust correlation of CO2 to global temperature.

        I’d be skeptical of such a result. In exactly the same way I am of the claim there ought to be such a result, or its absence means anything.

        In comparison, though flawed and needing work, the IPCC claims are relatively more plausible.

        And to me, unimportant overall.

        That CO2 is rising is the important issue. How it is going to matter is secondary, unless it goes to questions such as can this rise be slowed.

        CO2 rise is an external forcing in a spatiotemporal chaotic system, a perturbation of attractors, a stick bludgeoning a hornet nest that we do not understand, other than to understand we are dependent on the nest and the hornets in myriad and diverse ways.

        Pretty silly to carry on with the bludgeoning.

      • And.. Girma’s 60-year fixed period cuckoo’s egg simply doesn’t meet that criteria either.

        Where does this 60 years come from?

        Let us look what the TEAM says:


        1) The verification period, the biggest “miss” was an apparently very warm year in the late 19th century that we did not get right at all. This makes criticisms of the “antis” difficult to respond to (they have not yet risen to this level of sophistication, but they are “on the scent”).
        http://bit.ly/ggpyM1

        2) Here are some speculations on correcting SSTs to partly
        explain the 1940s warming blip.
        http://bit.ly/8SPNry

        3) We know the 2000s is also a warming blip.


        Conclusion:
        2000-1940=1940-1880=60!
        Global warming blips occur APPROXIMATELY every 60 years.

      • Girma

        If it isn’t a fixed period, but only an approximate one, then again a sine curve is inappropriate for the simple reason that a sine curve is based on a fixed — not “APPROXIMATE” — period.

        Read Mandelbrot for some fractal solutions that have an alternating high-low bias; many are quite suitable for this sort of curve, and don’t bear the connotation of a fixed period.

        Perhaps to some, the distinction is lost: why ought it matter that the period of sine curves be fixed rather than “approximate”, when after all the dependent variable is “approximate” (within error bars, or due correlation coefficients) too?

        Perhaps the simplest way to explain why this is wrong I have at hand is the shorthand of the ‘degrees of freedom’ concept.

        With one degree of freedom in two dimensions, a function has a correlation coefficient that meaningfully compares how well independent variables predict dependent variables through the function or hypothesis.

        With two degrees of freedom in two dimensions, that meaning collapses into two possibilities ambiguously, either:
        a) the correlation reflects goodness of prediction; or,
        b) the correlation reflects coincidence of independent variable.

        So, please, please, please stop using your sine curve because a few points hand fit. It’s an ugly abuse of the process of graphical analysis well outside the confines of the beautiful methods of that mathematical endeavor.

        The world is full of alternating functions. Find one of those if you must propose a curve, and find a mechanism that it explains.

        I don’t go around proposing a game can be called “chess” where all pieces move based on the roll of a dice, do I?

  72. Bart made a remark about me believing my own predictions. Pekka observed that he had not seen evidence of chaotic bifurcation in climate. The issues are interconnected so let’s look at both.

    We are in a specific state of the interdecadal Pacific Oscillation (IPO – and I do wish they would stop calling these things oscillations). I just googled this and got 30,800 hits – so it is far from unknown. These states tend to last for 20 to 40 in the instrumental and proxy record. A cool La Niña phase of the IPO sees +ve PDO and more frequent and intense La Niña. A warm El Niño phase sees more frequent and intense El Niño. The phases have an effect on global hydrology and on global surface temperature. As we are in a cool phase since 1998 and the phases last 20 to 40 years – it is a simple matter to connect the dots.

    Here is a study from PNAS – http://www.pnas.org/content/101/12/4136.full – that uses the PDO as a proxy of the Pacific variability and comes to a similar conclusion. Although they describe it as lees steep warming rather than my more bold claims of planetary cooling. I note the April UAH anomaly for March is -0.10 degrees C. and still heading down as a result of the current large La Niña – we shall see.

    Let’s have a look at the multivariate ENSO index of Claus Wolter. Open it up in a new window.

    http://www.esrl.noaa.gov/psd/people/klaus.wolter/MEI/

    There are a couple of things that we can see immediately. The period to 1976 was dominated by a La Niña and the period from 1977 was dominated by an El Niño. The shift in 1976/1977 is something known as the Great Pacific Climate Shift (2,300,000 hits). It is obvious in a range of records as an abrupt and nonlinear change. It was identified by Tsonis and colleagues in his Northern Hemisphere network model – which included both the PDO and ENSO – as chaotic. The other thing to be seen is the large 1998/2001 fluctuation in ENSO in a period identified in the network model as a phase transition. This clearly fits the definition of a Dragon King.

    The abrupt shifts are clearly seen in biological responses to varying upwelling of nutrient rich water– changes in abundance of Chinook salmon in North American stream, phytoplankton in the central Pacific and sardines in Monterey Bay amongst many others. This – along with abrupt changes in hydrological regimes – is the clearest evidence for abrupt change on a multi-decadal scale. That the surface temperature trajectories reflect these changes in Pacific states is only to be expected.

    Abruptness and non-linearity are signatures of complex dynamical systems and this provides clues about how Earth systems should behave and what to look out for – slowing down and noisy bifurcation followed by a settling into a new pattern. Predictability as we know it goes out the window.

    • Chief Hydrologist

      Why don’t you include the following graph in your description of “Great Pacific Climate Shift”?

      http://bit.ly/ePQnJj

      A picture is worth thousand words!

    • Chief

      Put another way: http://www.youtube.com/watch?v=4n5AfHYST6E

      :D

      Hope you find it novel.

    • Funny, I Googled “Great Pacific Climate Shift” and got only 5,390 hits.
      Of course I get many more if I do not use double quotes.

    • Rob,
      My point is that one observation tells about the thing observed, not about other things.

      Another point is that empirical separation of a bifurcation from other shifts requires much more information than we have.

      Similarly separating true chaos from other fluctuations requires much more information than we have available.

      We have many kind of evidence on the climate and its dynamics, but they do not prove any general picture. Thus the issue of practical importance is, how much we have learned to help us make projections of future climate under some specified external conditions and concentrating on some specific climate indicators. There is evidence that climate scientists have learned something of significance. There is also evidence that future climate is difficult to predict. Having an open mind means that all this evidence is taken into account and judgment made on its significance. Then the conclusions and uncertainties are described as well as possible to those who wish to have such information for their decision making.

      The result is not determined by claims on the existence of chaos in atmospheric or ocean processes as fundamental claims. The results are of a more limited nature, and the conclusions may be same for two alternative models of climate, one with real chaos and the other without, as the uncertainties related to any concrete need may be the same for a suitable pair of models (with chaos and with non-chaotic uncertainties). The concrete case dependent uncertainties are important, labels put on models are not.

      • Pekka,

        Where we have processes that occur over space and time with multiple feedbacks and multiple interactions – and this results in non-linear oscillations – we have a spatio-temporal chaotic system. This is not mere labelling but an important recognition of underlying physical processes.

        We know that the numerical models of climate are examples of temporal chaotic systems. Indeed it was this property of models that first alerted people to deterministic dynamical complexity, as opposed to randomness, in weather. Climate is widely recognised as chaotic – and the difference with weather is a mere matter of scale. ‘The global coupled atmosphere–ocean–land–cryosphere system exhibits a wide range of physical and dynamical phenomena with associated physical, biological, and chemical feedbacks that collectively result in a continuum of temporal and spatial variability. The traditional boundaries between weather and climate are, therefore, somewhat artificial.’ Hurrell et al 2009

        You respond adds nothing to the discussion by reference to acknowledged practitioners or literature. It is simply a subjective rant in a superficial idiom of objective science – while assuming a mantle of gravitas – signifying nothing.

        Seriously Pekka – I regard you as little better than any old partisan of the climate wars. You gave negligible understanding of the issue and wish only to defend a position by means of spreading confusion.

        Cheers

      • Rob,
        In some sense our positions are mirror images. You say that I try to spread confusion, when my own view is that I attack claims made to spread confusion.

        You select results that can be interpreted to tell that the chaotic features of climate models make them incapable of providing any useful knowledge.

        I admit fully that the models contain chaotic features, but my claim is that they do not necessarily make the models worthless for gaining knowledge on the consequences of adding CO2 to the atmosphere.

        I’m not ready to disregard as worthless the views of the numerous climate modelers who claim that the models provide essential foresight. This is my position in spite of the fact that I have my own thoughts on, how all modelers may fail due to common reasons and how their majority may be influenced strongly by the same biasing forces.

        When the knowledge is seriously incomplete we all are try to generalize from too little information. You have good evidence for some weaknesses, but in my view not enough for the general conclusions that you present. Knowing that there are chaotic elements in the system does not tell, how dominating they are. Strong generalizations cannot be justified by the data alone, but they are mostly based on subjective judgments.

      • You should read the Hurrrl paper. Whether it is complexity of climate or complexity in models – the predictions are horribly inaccurate.

        As in a lack of warming for another decade at least – http://www.pnas.org/content/107/5/1833.short

        I judge the models implausible on the basis of describing fundamentals – ENSO, PDO, AMO etc – and in predicting even decadal temperature changes.

  73. There are a couple of things that we can see immediately. The period to 1976 was dominated by a La Niña (bias) and the period from 1977 was dominated by an El Niño (bias).

    Not sure why I took bias out. Perhaps it was the visions of a variety of bird life circling around my head.

  74. Thanks Steven for the link.
    http://bit.ly/kkOf7C

    Do global temperature trends over the last decade falsify climate predictions?

    Observations indicate that global temperature rise has slowed in the last decade (Fig. 2.8a). The least squares trend for January 1999 to December 2008 calculated from the HadCRUT3 dataset (Brohan et al. 2006) is +0.07±0.07°C decade–1—much less than the 0.18°C decade–1 recorded between 1979 and 2005 and the 0.2°C decade–1 expected in the next decade (IPCC; Solomon et al. 2007). This is despite a steady increase in radiative forcing as a result of human activities and has led some to question climate predictions of substantial twenty-first century warming (Lawson 2008; Carter 2008).

    El Niño–Southern Oscillation is a strong driver of interannual global mean temperature variations. ENSO and non-ENSO contributions can be separated by the method of Thompson et al. (2008) (Fig. 2.8a). The trend in the ENSO-related component for 1999–2008 is +0.08±0.07°C decade–1, fully accounting for the overall observed trend. The trend after removing ENSO (the “ENSO-adjusted” trend) is 0.00°±0.05°C decade–1, implying much greater disagreement with anticipated global temperature rise.

    We can place this apparent lack of warming in the context of natural climate fluctuations other than ENSO using twenty-first century simulations with the HadCM3 climate model (Gordon et al. 2000), which is typical of those used in the recent IPCC report (AR4; Solomon et al. 2007). Ensembles with different modifications to the physical parameters of the model (within known uncertainties) (Collins et al. 2006) are performed for several of the IPCC SRES emissions scenarios (Solomon et al. 2007). Ten of these simulations have a steady long-term rate of warming between 0.15° and 0.25ºC decade–1, close to the expected rate of 0.2ºC decade–1. ENSO-adjusted warming in the three surface temperature datasets over the last 2–25 yr continually lies within the 90% range of all similar-length ENSO-adjusted temperature changes in these simulations (Fig. 2.8b). Near-zero and even negative trends are common for intervals of a decade or less in the simulations, due to the model’s internal climate variability. The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate.

    The global warming rate from 1998 to 2008 given above was 0.08 deg C per decade.

    The trend after removing ENSO is 0 deg C per decade.

    The global warming rate from 1998 to 2010 still is 0.08 deg C per decade.
    http://bit.ly/mC64Hg

    It is now 13 years with zero trend.

    Two years are left for AGW.

    It is fate is not promising.

    • steven mosher

      I have an Idea for you. Why dont you write up a proper post, submit it to Lucia and folks at her place will give your ideas a proper thrashing.

      • Steven mosher

        folks at her place will give your ideas a proper thrashing.

        What I am saying others are saying in a published paper:

        Special Supplement to the Bulletin of the American Meteorological Society
        Vol. 90, No. 8, August 2009

        Observations indicate that global temperature rise has slowed in the last decade (Fig. 2.8a). The least squares trend for January 1999 to December 2008 calculated from the HadCRUT3 dataset (Brohan et al. 2006) is +0.07±0.07°C decade–1—much less than the 0.18°C decade–1 recorded between 1979 and 2005 and the 0.2°C decade–1 expected in the next decade (IPCC; Solomon et al. 2007). This is despite a steady increase in radiative forcing as a result of human activities and has led some to question climate predictions of substantial twenty-first century warming (Lawson 2008; Carter 2008).

        El Niño–Southern Oscillation is a strong driver of interannual global mean temperature variations. ENSO and non-ENSO contributions can be separated by the method of Thompson et al. (2008) (Fig. 2.8a). The trend in the ENSO-related component for 1999–2008 is +0.08±0.07°C decade–1, fully accounting for the overall observed trend. The trend after removing ENSO (the “ENSO-adjusted” trend) is 0.00°±0.05°C decade–1, implying much greater disagreement with anticipated global temperature rise.

        Steven, ask the Meteorological Society why they are saying “much greater disagreement with anticipated global temperature rise.”

        I have nothing to do with that as the society says exactly what I regularly say.

        What I don’t agree is with the society’s statement “we expect that warming will resume in the next few years”, as this is an unscientific statement.

      • Girma, can you pls provide the title and author of the paper, and a web link if possible. thx.

      • Judith

        Here are the details of the article and links.

        DO GLOBAL TEMPERATURE TRENDS OVER THE LAST DECADE FALSIFY CLIMATE PREDICTIONS?—J. KNIGHTHT, J. J. KEENNEDEDY, C. FOLLLLANDD, GG. HARRISS, S. JONESES, M. PALMELMER, D. PARKEKER, AA. SCCAIFEFE, ANDD P. STTOTTTT
        http://bit.ly/mvq4Rk

        The above article is also given in the following publication.

        STATE OF THE CLIMATE IN 2008
        Special Supplement to the Bulletin of the American Meteorological Society
        Vol. 90, No. 8, August 2009
        Pages 23 & 24
        http://bit.ly/kkOf7C

        The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate.

      • Sorry my cutting and pasting has some error. Here is a correction:

        Do global temperature trends over the last decade falsify climate predictions?—J. Knight, J. J. Kennedy, C. Folland, G. Harris, G. S. Jones, M. Palmer, D. Parker, A. Scaife, and P. Stott
        http://bit.ly/mvq4Rk

    • Nebuchadnezzar

      A few things:
      1) 0.08 is greater than zero. A trend of 0.08 would have to continue for longer than the stated 15 years to cause a discrepancy.
      2) you aren’t using the ENSO adjusted data set.
      3) you are ignoring GISS and NCDC data sets because HadCRUT3, having a lower trend, better suits your particular point of view.

      The 2-page article is also here (to save you downloading the whole BAMS state of the climate report):
      http://www.metoffice.gov.uk/media/pdf/j/j/global_temperatures_09.pdf

      • Observations suggest that the temperature rise is likely to be limited over the next decade. You should really be aware of this as an emerging theme in peer reviewed literature and the reasons behind it.

        http://www.pnas.org/content/107/5/1833.short

        The decadal changes are due to ocean variability. Moreover, it is misleading to ‘adjust’ for ENSO – even if that were ever more than a gross approximation.

      • Nebuchadnezzar

        Even if it is misleading, that was what was done in the Knight et al. paper. If Girma wants to rely on that quote then Girma needs to do the calculation right. Girma also needs to do the calculation for GISS and NCDC otherwise it’s just cherrypicking.

      • The datasets are not statistically different. You both refer to the same section of the BAMS report – whiuch essentially says that there has been negligible warming last decade but that 15 years is required to falsify the models.

        Now the models are deterministic complex dynamical systems with whose plausibility rests on 2 grounds. The a priori model formulation and a subjective judgement of a posteriori solution plausibility. The problems there are legion.

        Disaggregating ENSO is a thing that varies from years to year. What they said was that any the small amount of warming was due to ENSO – within fairly broad error bounds. It was in fact 2 things an energy decrease from solar irradiance in the 11 year cycle and a change in cloud cover that is associated with ENSO resulting in less reflected SW – seen in the CERES data.

        The next 10 years – as in the PNAS paper – is likely to see more intense and frequent La Nina and to be cooler – thus falsifying AGW in the minds of just about everyone. Get used to it.

      • Nebuchadnezzar

        Figure 2.8 (b) in the article shows the trends from the three different data sets. Over the crucial period, they are quite different. GISS and NCDC have higher trends and fall in a different, ‘more likely’ part of the model distribution. In that sense, statistically they are different. In what sense are they statistically the same?

        What they said was that any the small amount of warming was due to ENSO
        At the time, people were saying that the “slow down” could be explained purely in terms of ENSO variations. This paper explicitly removed that component and still found a slow down. That was an important possibility to try and discount. However, other people have done similar analyses and found that the ENSO component is quite different. It’s clearly not an easy thing to do and it may be too simplistic a way of thinking about things.

        One alternative reason for the slowdown mentioned in the article is internal variability. Solar changes and data biases are also offered as potential contributors. Aersols (manmade particularly over Asia and volcanic) have been suggested elsewhere as has the possibility of sequestration of heat into the deep ocean. You have other ideas and yet more have been advanced elsewhere. We’re not short of hypotheses.

        These different suggestions aren’t going to be resolved just by looking at short trends in global temperature. As you say, we need to look at a range of different data sets and variables, evaluate a range of hypotheses and allow in our deliberations for the uncertainties in instruments that may or may not have the accuracy, stability or precision required to answer the questions posed of them. On top of that the chances of it having a single cause are approximately zero.

        One interesting thing, that hasn’t been looked into is how the coverage of the different data sets (HadCRUT, GISS and NCDC) affect the comparison with the models. HadCRUT has large areas of missing data over the poles and a less complete land analysis than ocean analysis at lower latitudes. GISS and NCDC perform different amounts of datafilling, so they’re not measuring exactly the same thing and none of the observed analyses is measuring precisely what the models produce which is globally complete 2m air temperatures.

        I think a direct comparison over the area that is common to all these analyses will help resolve some confusions because it removes one of the big uncertainties in observed estimates of global temperature and that is what is happening over the poles.

        If I had to guess what the result would be, I would say that the north pole is warming faster than the models ‘expected’ (it would be in the upper half of a distribution like that in the article but for the Arctic alone) but that over the HadCRUT area with all its gaps, the discrepancy between models and obs would be very similar for the three observed datasets and it would be slightly worse than is suggested in Knight et al. or, for that matter, in the Knappenberger et al. Heartland presentation.

        The PNAS article is interesting. It’s not the first paper to suggest that the initialisaiton of natural internal cycles of variability is important in decadal forecasting, nor the first to make a decadal prediction. Do you think this one is more likely to come out right?

        Right now, I’m wary of anyone who says they definitely know exactly what is going on. My philosophy is that we will see what the next 10 years brings and whatever happens, we will all have to get used to it in one way or another.

      • I’d suggest that the differences between all 5 temperature datasets are all insignificant unless you neglect limits of error. http://www.appinsys.com/GlobalWarming/GW_Part2_GlobalTempMeasure.htm#comparison

        ‘The least squares trend for January 1999 to December 2008
        calculated from the HadCRUT3 dataset (Brohan et al. 2006) is
        +0.07±0.07°C decade –1—much less than the 0.18°C decade
        –1 recorded between 1979 and 2005’

        ‘ The trend in the ENSOrelated component for 1999–2008
        is +0.08±0.07°C decade–1, fully accounting for the overall observed trend. The trend after removing ENSO (the “ENSO-adjusted” trend) is 0.00°±0.05°C decade–1, implying much greater disagreement with anticipated global temperature rise.’

        Contrary to your interpretation – what this analysis found was that ENSO fully accounted for all of the small rise. The internal variability in other words was most obviously ENSO related – which is after all the source of most global climate variability on interannular to decadal scales.

        Here is another ‘interesting’ PNAS paper on models this time – http://www.pnas.org/content/104/21/8709.full

        Keenlyside et al 2007 and Tsonis and colleagues 2007 and 2009 – suggested a decadal slow down that involved ocean states – especially in the Pacific. You need to understand these states to understand why they are all equivalent and why they are likely to be correct. You need to understand the source of ‘internal climate variation’ to make any sense of temperature variations and not simply parrot a narrative in a superficial idiom of science.

        A cool phase of the Interdecadal Pacific Oscillation is with us. These last 20 to 40 years. The cool phase involves more frequent and more intense La Nina. La Nina is associated cooler temps with more low level cloud as cloud forms over cooler SST. Connect the dots.

        These are not cycles or oscillations as such – they are chaotic bifurcations. ‘The global coupled atmosphere–ocean–land–cryosphere system exhibits a wide range of physical and dynamical phenomena with associated physical, biological, and chemical feedbacks that collectively result in a continuum of temporal and spatial variability. The traditional boundaries between weather and climate are, therefore, somewhat artificial.’ Hurrell et al 2009 – http://www.gfdl.noaa.gov/bibliography/related_files/Hurrell_2009BAMS2752.pdf

      • Nebuchadnezzar

        You are mixing different quantities by comparing tropospheric temperatures and surface temperatures. They’re not the same quantity and they can’t be used interchangeably. It’s not an apples to apples comparison.

        The differences between the three *surface* temperature data sets have a number of causes some of which are explicable in terms of data coverage. The global average from GISS warms faster than HadCRUT3 over the most recent decade because GISS extrapolate over the Arctic. If you plot the average over the common area, the agreement is much tighter.

        Your opinion on this might differ, but if you think the spread is insignificant simply due to limits of error, you still can’t pick the one you like. You have to consider the full range. To borrow your own phrase, don’t “simply parrot a narrative in a superficial idiom of science”.

        Contrary to your interpretation – what this analysis found was that ENSO fully accounted for all of the small rise. The internal variability in other words was most obviously ENSO related

        That’s not contrary to my opinion. It’s more or less a restating of it, added to some confusion about what we’re each referring to as far as I can tell.

        I agree that in order to understand the current climate, we need to know which “internal variations” are at work. The reason I wasn’t more specific is that the diagram (2.8b) from Knight et al. doesn’t differentiate between different modes of variability. It’s worth noting that their final figure (2.8c) shows that perhaps the Southern Oceans might have played a part in the slow down.

        Regarding your prediction of cooling, I don’t know whether it’s right or wrong. I’m happy to wait and see with the proviso that since you only think it likely, we might never know if you were right or wrong. I was intrigued as to why you particularly favoured one prediction, so thanks for the explanation.

      • you ‘At the time, people were saying that the “slow down” could be explained purely in terms of ENSO variations. This paper explicitly removed that component and still found a slow down.’

        I think you clearly show bad faith in first of all claiming one thing.

        me “Contrary to your interpretation – what this analysis found was that ENSO fully accounted for all of the small rise. The ‘internal variability’ in other words was most obviously ENSO related”

        ‘That’s not contrary to my opinion. It’s more or less a restating of it, added to some confusion about what we’re each referring to as far as I can tell.’

        And then claiming something else entirely.

        You misrepresent me entirely – I don’t predict anything. As I say – we are in a cool mode of the IPO.

        The comparison of temperatures was based on normalised (by Professor Ole Humlum) values if you cared to check.

        Figure 2 ‘(b) ENSO-adjusted global mean temperature changes to 2008 as a function of starting year for HadCRUt3, GiSS dataset (Hansen et al. 2001) and the NCDC dataset’ (Smith et al. 2008) – but really nothing to do with any other type of variation at all. I don’t know what you are claiming here at all.

        We might have had cooler global surface temps in the sub Antarctic Southern Ocean as a cause of cooling? This is consistent with a more negative SAM index since the late 1990’s – but you extrapolate from one plot to a broad suggestion. One that is on track but needs a hell of a lot more detail to mean much at all – except to add to a general air of misdirection.

        As I say – I think you are arguing in bad faith. A pointless exercise at any time.

      • Nebuchadnezzar

        “You misrepresent me entirely – I don’t predict anything. As I say – we are in a cool mode of the IPO.”

        My bad then. I was thinking of these statements of yours which sounded to me like predictions:

        “A cool phase of the Interdecadal Pacific Oscillation is with us. These last 20 to 40 years. The cool phase involves more frequent and more intense La Nina. La Nina is associated cooler temps with more low level cloud as cloud forms over cooler SST. Connect the dots.”

        and from earlier:

        “The next 10 years – as in the PNAS paper – is likely to see more intense and frequent La Nina and to be cooler – thus falsifying AGW in the minds of just about everyone. Get used to it.”

        I’m trying to work out what you mean and suggesting other things that might also be going on. It’s not intended as misdirection or bad faith so please don’t take it in that spirit.

      • ‘Verdon and Franks (2006) used ‘proxy climate records derived from paleoclimate data to investigate the long-term behaviour of the PDO and ENSO. During the past 400 years, climate shifts associated with changes in the PDO are shown to have occurred with a similar frequency to those documented in the 20th Century. Importantly, phase changes in the PDO have a propensity to coincide with changes in the relative frequency of ENSO events, where the positive phase of the PDO is associated with an enhanced frequency of El Niño events, while the negative phase is shown to be more favourable for the development of La Niña events.’

        See for instance the decadal variability in ENSO in Claus Wolter’s MEI – http://www.esrl.noaa.gov/psd/people/klaus.wolter/MEI/

        ENSO is the biggest cause of global surface temperature variability. For instance – more than 1 degree C between 1998 and 2001 in the monthly UAH data.

        What I am describing is the current state of the Interdecadal Pacific Oscillation – and not a prediction at all.
        As in the papers I referenced – the IPO is likely to persist in a cool phase for another decade at least.

        Now that should be quite simple – even if I need to repeat it three times to you. You suggest as I said clearly – nothing of any substance.

      • Nebuchadnezzar

        “What I am describing is the current state of the Interdecadal Pacific Oscillation – and not a prediction at all. As in the papers I referenced – the IPO is likely to persist in a cool phase for another decade at least. ”

        I’m willing to accept that you think that a statement about the likely path of the future is not a prediction. You know your own mind and intentions better, I hope, than anyone.

        However, do you accept that someone reading your words might, without any intention of misrepresenting you, have formed a contrary opinion? Saying that the IPO is likely to persist in a cool phase for another decade at least seems to me to be a statement about the future course of the climate and hence, to my mind, a prediction.

      • You are very persistent. We are in a cool phase of the Interdecadal Pacific Oscillation. As the name suggests – these tend to persist for decades with more intense and frequent La Nina. La Nina cool the planet – as a result of cloud radiative forcing as one factor. La Nina of course results from cold water upwelling in the eastern Pacific and spreading westward. Low level marine stratiform cloud is negatively correlated with sea surface temperature. It gives secular changes on decadal timescales in cloud across the Pacific observed both from the surface and from satellites.

        Cloud was always recognised as a major area of weakness of
        climate theory – it is what it is.

      • Since 1998, for 13 years, the global mean temperature trend has been flat (0.01 deg C warming per decade: http://bit.ly/m6hGhZ) at about 0.4 deg C.

        GMTA for 2011 has started from the low value of 0.21 for Jan, 0.26 for Feb & 0.32 for March. As a result, it is likely that the GMTA for 2011 will be below the average since 1998, further continuing the cooling trend.

        Only two years left to verify the 15-year criterion.

        The ENSO adjusted data gives even lower warming rate.

        HADCRUT3 is the data used by the IPCC.

      • Nebuchadnezzar

        It’s interesting to replot your diagram using GISSTEMP instead of HadCRUT3:
        http://www.woodfortrees.org/plot/gistemp/from:1998/compress:12/plot/gistemp/from:1998/trend

        They don’t seem to have the NCDC data.

        The trends is, relatively speaking, much higher for GISS and, while GISS and NCDC might not have been used in the IPCC, they were used in the paper you were quoting and they were also ENSO adjusted in that paper. You can’t take the quote and apply it to the non-adjusted series.

        Do you have a plot showing the ENSO adjusted warming rate?

      • Do you have a plot showing the ENSO adjusted warming rate?


        ENSO and non-ENSO contributions can be separated by the method of Thompson et al. (2008) (Fig. 2.8a). The trend in the ENSO-related component for 1999–2008 is +0.08±0.07°C decade–1, fully accounting for the overall observed trend. The trend after removing ENSO (the “ENSO-adjusted” trend) is 0.00°±0.05°C decade–1, implying much greater disagreement with anticipated global temperature rise.

        According to this info, subtract 0.08 deg C per decade to remove the effect of ENSO

      • Nebuchadnezzar

        Yes, that’s the series to 2008, but you are making claims about data up to the present and the ENSO component and its trend will depend on the period. The ENSO component isn’t just a straight line and we’ve had a moderately strong La Nina event and a moderate El Nino since 2008. It’s better to do the calculation than just guess.

      • Accepted

  75. The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate.
    http://bit.ly/kkOf7C

    That is the first statement I read to verify AGW.

    That is science.

    • Which only means you can see one such 15+ year trend out of every 20 disjoint 15 year periods and be perfectly within the prediction.

      Indeed, you can see two such periods out of 20, and only have a fractional likelihood you can reject the prediction.

      Or three such periods, for a larger fractional likelihood of rejection.

      Statistics is not an on-off switch.

      Try doing the statistics not on the trends, but on the amplitudes. It’s far simpler, and one is less suspicious of trend analyses after the more fundamental analyses of data are completed.

      Or, since Chief has laid out extensive and very impressive discussions, speak to them. They’re really interesting stuff.

  76. 1. Do check out Terence C Mills’ paper in Journal of Cosmology (In press): Is Global Warming Real? Analysis of Structural Time Series Models of Global and Hemispheric Temperatures.
    http://journalofcosmology.com/ClimateChange112.html

    Mills concludes “..the optimal forecast of the long term trend is the current estimate of the trend, thus ruling out continued global warming, and, indeed, global cooling”.

    2. The whole Hempel stuff here is a red herring. All met stations in the USA show zero statistically significant warming between 1960 and 2006. That is a lot of white ravens/black swans.

    • Heh, neither random walk nor linear trend. We are cooling, folks; for how long even kim doesn’t know.
      =========

  77. Bart, the 15 year time period only talks to the hypothesis as presented by the models and does not address the hypothesis of AGW. Many skeptics, myself included, would argue that 30 years is too short a period of time to evaluate the climate especially when you have a positive PDO and a positive AMO and, if you want to argue for a small transient portion to the climate sensitivity, a positive solar component.

    Nebuchadnezzar, an underestimate of polar amplification or an underestimate of the warming caused by a positive AMO? That will be difficult to seperate until we have a negative AMO.

    • steven

      I agree. But I have a quibble reflex.

      Myself, I favor going with the longest time period that allows meaningful analyses, case by case.

  78. Maybe complex analysis of odds of a black swan are unnecessary and counter-productive, victims of the Berra Observation. Economics and financial analysis may have been chasing their mathematical tails, for example. Simple rules that balance and minimize risks match or out-perform sophisticated computer models.

  79. The whole black swan concept is idiotic in this context. We are basically gambling on something totally unexpected and undefined to turn up that will make everything OK. But by definition, it will most probably not turn up. Even then, if it does turn up, why would it make everything OK? Since we don’t know what it does, it will just as likely make AGW much worse.

    • Not at all. Since AGW is negligible, and beneficent if real, the “black swan” would be some highly improbable serious negative effect. It may be important to keep a “weather eye out” for such an eventuality, but it’s hardly worth dedicating the entire governance of the planet to it!

  80. Hello there, You’ve done a great job. I will definitely digg it and personally suggest to my friends.

    I’m sure they will be benefited from this web
    site.