RealClimate logo

The uncertainty prayer

Filed under: — gavin @ 29 June 2010

Seen at a meeting yesterday:

Grant us…
The ability to reduce the uncertainties we can;
The willingness to work with the uncertainties we cannot;
And the scientific knowledge to know the difference.

(Drawn from a white paper on the use of climate models for water managers).


140 Responses to “The uncertainty prayer”

  1. 101
    Ric Merritt says:

    Septic Matthew @ 1 July 2010 at 12:08 PM (#79):

    Care to make a substantial public bet under your real full name on the trend in sea ice over the next 3 decades? I’d be happy to take the “declining” side for total sea ice, both arctic and antarctic, if you want the “increasing” side.

  2. 102
    Ray Ladbury says:

    Andreas, are you afraid of math?

  3. 103
    Andreas Bjurström says:

    Yes, annoyance are often a problem, especially in inter-cultural dialogue.

    I agree with your ‘boxes’ argument. In addition to that it is important with a bit of history, to see where things are coming from. One is of course not post-modern on the merit of being specific about something.

    Post-modernism is a reaction to modernism (a complex box). Modernism includes rationality, objectivity, the search for truth, laws of nature, progress, and many other things. Post-modernism is sceptical towards that. I tried to specificly adress one of these things, post-modern inspired research as a reaction to the scientific enterprice of generalising to find laws of nature and the like. But lets drop this sidetrack.

    Yes, we can learn some things from acid rain or ozone depletion issues. For example, that these issues was contested by “sceptics” for the same reasons that climate change is contested by sceptics. We can also learn that uncertaintly was not critical to solve previous environmental problems. The DDT problem for example was solved despite great uncertaintly in the limited scientific literature. Why so? Because there was an easy technical solution to the problem. Climate change is different …

  4. 104
    Andreas Bjurström says:

    102 Ray Ladbury,
    No, I´m not afraid of math (I was good at math in my youth, and I do use statistical analysis in my research, especially MDS).

    I just acknowledge the limits of math and that relying to heavily on math can in fact be a great risk.

  5. 105
    Anonymous Coward says:

    Perhaps Andreas Bjurström would like to link a short piece he wrote outlining his case about the “linear model”, the IPCC and related issues. I can’t figure out what his point is, though I tried. I think the low-probability risks are not dealt with seriously enough considering the weight they would likely have in a full risk assessment but that’s probably not his point.

    I’m surprised he tries to blame anyone but social scientists themselves for their infatuation with numbers and (poorly understood) statistics. The humanities have largely resisted this trend and I don’t think they’re being disrepected as a result.

  6. 106

    My impression of what Andreas Bjurström was trying to say was that it was valid, but not very well put. Here is an example of how relying on rational argument is not going to solve the problem of climate change.

    It is an article from the magazine “Planet Earth” produced by the UK’s Natural Environment Research Council (NERC).

    You, your car, and climate change


    Cheers, Alastair.

  7. 107
    Andreas Bjurström says:

    105 Anonymous Coward,

    My point was something like: If reduced scientific uncertainty doesn´t make us more certain about what to do, than why is uncertainty so important? perhaps it is important primarily for participants of some scientific disciplines, e.g. it is their professional interest, not our interest. Moreover, the uncertainty regarding physical aspects are rather low (the issue all seems to be concerned with, including sceptics) whereas the uncertainty in biological systems and social systems is far greater. Why is the small uncertainty of the physical world so important when we successfully handle far greater uncertainties every day? (I get a feeling Im changing my point every time, but this was a try, and I honestly dont want to take up to much space here, but many people always reply to what I say here for some strange reason).

    Try Roger Pielke Jr blog, he post something about the linear model etc. every other day.

  8. 108
    Septic Matthew says:

    101, Ric Merritt: Care to make a substantial public bet under your real full name on the trend in sea ice over the next 3 decades? I’d be happy to take the “declining” side for total sea ice, both arctic and antarctic, if you want the “increasing” side.

    No. I claim no knowledge about future climate, other than that weather will fluctuate.

    Of the bets available to me, I would rather bet that in 10 years time renewable alternative energy sources will be commercially competitive against petroleum without subsidies. I do not use enough electricity off the grid to justify purchasing a PV system, but I expect that within 10 years that will change; I review the math about 2 times per year to see if it is yet time to place the bet on a home PV system. If I were younger (I’m 63) I’d be investing in renewable energy companies.

    I’d also discuss a bet about nuclear power, but nuclear has been declared off-topic.

  9. 109
    Ray Ladbury says:

    Andreas, did you not read what I wrote? I said that the appropriate action is much more severe if, for example, CO2 sensitivitity is 4.5 or higher than it is if its between 2 and 3. The uncertainty makes it impossible at present to bound risk. Bounding risk is a prerequisite for policy and resource allocation for mitigation.

    The bounds we are able to establish at present would demand we slam the brakes on fossil fuel consumption immediately.

  10. 110
    David B. Benson says:

    Septic Matthew (108) — Go to
    for that topic.

  11. 111
    mike roddy says:

    The main point about uncertainty is that predictions by their nature are uncertain, and make many non scientists or less thoughtful people uneasy. When climate scientists express quite openly that the degree of future warming is uncertain, that is pounced on my the lazy or conflicted as an excuse to discard these predictions altogether. Many place more credibility in a preacher who tells them that they will go to heaven (or hell)- it’s the faux authority, not the content.

    Dealing with predictive imprecision is second nature to scientists, but is the opposite for much of the public. This is an overlooked gap in communications between scientists and the public, and if it’s crossed a lot of things might become possible.

  12. 112
    Harmen says:

    But even if we take the skewness risk of climate change into account…
    I think we are still underestimating the risk of rising CO2 concentrations..

    Climate change might have consequences that nobody has considered yet.

    Furthermore there might be consequences of rising greenhouse gas concentrations beyond climate change and ocean acidification that nobody has yet considered.

    Both could have an extreme impact.

    This is in line with the black swan theory of Taleb..

    “Writing in the New York Times, Taleb asserted:

    What we call here a Black Swan (and capitalize it) is an event with the following three attributes. First, it is an outlier, as it lies outside the realm of regular expectations, because nothing in the past can convincingly point to its possibility. Second, it carries an extreme impact. Third, in spite of its outlier status, human nature makes us concoct explanations for its occurrence after the fact, making it explainable and predictable. I stop and summarize the triplet: rarity, extreme impact, and retrospective (though not prospective) predictability.”

  13. 113
    Hank Roberts says:

    > … nothing in the past can convincingly point to its possibility.
    > Second, it carries an extreme impact …

    Harmen, the reason climate scientists study natural events in the past is to get a good idea what the many possible forcings are. Deep time is really astonishingly deep, and the paleo record is still accumulating.

    But to worry about a ‘black swan attack’ event that has _ever_ occurred before, you need to imagine _two_ of them; one to subtract the forcing we now attribute to carbon dioxide based on known physics, and another to produce a comparable forcing out of some mysterious gray area unknown to science.

    Have a look at the evolution ‘debates’ for the uses of the ‘gray area’ or ‘gaps’ argument to suggest some vast unknown force behind the curtain of time.

    Often it’s just used to try to divert people from the known facts, e.g.:

  14. 114
  15. 115
    Hank Roberts says:

    Speaking of uncertainty, recommended:

    “… when I meet a conservative AGW skeptic who says all that (and I have), I am all kisses and flowers. And so will be all the atmospheric scientists I know. That kind of statement is logical, patriotic and worthy of respect. It deserves eye-to-eye answers.

    But alas, such genuine “skeptics” are rare.”

  16. 116
    Ray Ladbury says:

    Harmen, what you are discussing is my continual reality in my day job. I am tasked to achieve multiple “nines” of reliability even while haveing test samples in the single (and sometimes low single) digits. Of course we cannot do this without having a model, and of course the model cannot be trained on completely representative data. It is the sort of problem that keeps a decent risk mitigation expert up at nights–all the more so when it involves the fate of human civilization.

    I agree that past events cannot be guarantors of future behavior. However, we have a few billion years of Earth history to develop at least some sort of Bayesian Prior.

  17. 117
    Hank Roberts says:

    PS, Brin makes an excellent comparison there to the tobacco industry delay, quoting some analyst:

    “… with enough smokescreens raised to delay public acceptance, there is far more time to gradually unload stock, and perhaps even reposition the companies in the most vulnerable industries.”

    That’s a key observation about manufactured uncertainty.

    Note that since Brin wrote that in March (Skeptic Magazine and his blog) and at Salon in February (, the Supreme Court has let tobacco off the hook for damages, leaving only the wrist-slap for lying for decades. This is a precedent that may let oil and coal — as tobacco did — stop the denial, admit they lied, and ‘move on’ to higher stock prices.

    Denial for industries is a political-judicial-legislative-marketing tactic, not a philosophical position.

    We may see much less funding for ‘advocacy science’ now that the Supreme Court has established a precedent, because providing denial as grist for advocacy (legal argument) is now not going to be needed at least in the US. Perhaps other countries and the EU will see more of it though.

  18. 118
    Jimbo says:

    Comment by t_p_hamilton — 1 July 2010 @ 10:29 PM
    You were the one who mentioned pressure, not Mike or Gavin in his response to Mike.

    RESPONSE: I know I was the one who brought up pressure and that was my whole point. Instead we got a simpler response from Gaving which did not inclued atmospheric pressure while mentioning Co2 and composition. What’s wrong with a simple obervation? Furthermore, could someone let me know how much of an effect, temperature wise, Venus’ very proximity to the Sun has on temperatures there>

    Anyway, while on one of NASA’s websites I read something interesting about Venus.
    Venus Express has certainly confirmed that the planet has lost a large quantity of water into space.
    “It happens because ultraviolet radiation from the Sun streams into Venus’s atmosphere and breaks up the water molecules into atoms: two hydrogens and one oxygen. These then escape to space.”

    (Please not that this quote is not intended to relate to AGW – just something I found interesting – snip at will!)

  19. 119
    Harmen says:


    >>Often it’s just used to try to divert people from the known facts<<

    I am certainly not trying to distract from the facts.I subscribe to the conclusions of the IPCC reports but if we only focus on the facts then i think we will do to little about rising GHG concentrations.

    The IPCC tends to focus on effects are likely or very likely. That is good from a scientific perspective..

    But when we talk optimal policy (optimal carbon cap)and decision making under uncertainty i think we also have to make risk assesments..

    i.e. we also have to think about low probability consequences of carbon emissions that could have an extreme impact.

    Let me try to explain by an example…

    Consider the possibility there is a link between temperature and earthquakes/tsunamis..

    Global warming may bring tsunami and quakes: scientists
    By Richard Meares
    LONDON | Wed Sep 16, 2009 5:07pm EDT
    Speakers were careful to point out that many findings still amounted only to hypotheses, but said evidence appeared to be mounting that the world could be in for shocks on a vast scale.

    It lies outside the realm of regular expectations? extreme impact?

    Should this risk influence the optimal carbon cap?
    I think it should..

  20. 120
    Anonymous Coward says:

    OK, I checked Pielke Jr.’s blog and found an entry on the “linear model”. It’s science -> public opinion -> policy. Some commenters here seem to think according to something like this model. But this is clearly not what the IPCC is about (the I stands for intergovernmental!). Who actually works according to this model? I guess this is what Hansen might be doing (some of the time anyway). Who else?

    Andreas, the way you use of general pronouns like “we” makes it impossible for me to understand what you’re trying to say. If you’re a social scientist, then you must know how to describe what goes on in societies more effectively.
    Yes, there are uncertainties which mainly concern the scientists. But there are physical uncertainties with a broader relevance. Best not generalize so much if you don’t have anything to say which would apply to all uncertainties. Surely you’re not suggesting that, say, ice sheet dynamics are irrelevant to the huge public works which would protect low-lying areas against flooding.

  21. 121
    Rod B says:

    Hank (115), I was quite impressed with your referenced Brin’s commentary on skeptics vs deniers. He went off the track a couple of times IMO by doing some of the negative things he was describing like that old chestnut of the tobacco-climate cabal, and the petro-media conspiracy. But I would give him some leeway. I thought it very astute and insightful.

  22. 122
    Radge Havers says:

    “Any fool can criticize, condemn and complain and most fools do.”
    Benjamin Franklin

    Get automated “critiques” at The Postmodernism Generator. (Hit refresh for additional blasts of foolishness.)

  23. 123
    Hank Roberts says:

    Rod, compare your writing to Brin’s criteria for honest skeptics.
    Look at the facts, not labels from the PR operations. That’s his point.

  24. 124
    Ray Ladbury says:

    While I agree that we should be cannot ignore low-probability, high-consequence events, we already have the problem that risk cannot be bounded even if we confine ourselves to increased drought, flooding due to impulsive rain events and the impact of variable climate on agriculture. This leaves us comparing a series of unbounded risks–hardly ideal from a mitigation perspective.

    If this were a satellite project, the appropriate approach would be to slam on the breaks and try to refine the risks less conservatively while working out mitigations. Or put another way, risk avoidance is the only viable mitigation approach under such circumstances.

  25. 125
    Ray Ladbury says:

    Jimbo, never apologize for being interested in science. Venus is a fascinating planet–although I don’t think it will become a vacation destination any time soon.

  26. 126
    Thomas says:

    Jimbo, I’m probably not the best person to answer your question, but until the big guns deem it necessary to respond I’l through out a few tidbits:

    Of course the greater insolation at Venus’ distance from the sun does matter. It is interesting however that the total absorbed solar radiation per meter squared on Venus is less than on the earth. This is a consequence of the high albedo from the dense cloud cover. To risk being off by a few hundredths earths albedo is roughly 30%, and Venus’ is roughly 70%. So total solar absorbed radiation is actually less on Venus. But, I’m pretty certain that the higher solar constant drove the processes which created the runaway greenhouse effect. But once you get that strong a greenhouse effect, higher cloudbased albedo just doesn’t help enough.

    The loss of water is interesting. It requires enough water of hydrogen to get high into the planets exosphere, so that the low molecular weight/ high velocity hydrogen can escape. The fact that Venus’ magnetic field provides little protection from the solar wind, (I think) also increases the loss mechanism. It should be noted that young solar type stars, because of their likely much faster spin rates have very much stronger solar winds (perhaps a hundred times).

  27. 127
    Harmen says:


    Some of these risk are difficult to quantify but they are significant because they could have an extreme impact.

    My plea is that we should not to ignore these risk even if we don’t fully understand them. If we can’t quantify, we have to add qualitative assesments.

    And when it comes mitigation..I think we are in agreement. i do belief we have to “slam on the breaks” because of the risks involved.

  28. 128
    jyyh says:

    with an extremely dry ‘hah’, i note black swans may be created in the gulf of mexico in the autumn.

  29. 129

    Jimbo 118: could someone let me know how much of an effect, temperature wise, Venus’ very proximity to the Sun has on temperatures there

    BPL: Venus has a solar constant of 2,611 watts per square meter and a bolometric Russell-Bond spherical albedo of 0.750. Therefore the absorbed climate flux is

    (2611 / 4) (1 – .75) = 163 watts per square meter

    Its radiative equilibrium temperature is therefore

    (163 / sigma)^0.25

    where sigma is the Stefan-Boltzmann constant, 5.6704 x 10^-8 watts per square meter per Kelvin to the fourth in the SI. So Te = 232 K.

    Compare to Te = 255 K for the Earth. Venus is closer to the sun, but its high reflectivity means it absorbs LESS sunlight than the Earth, and all else being equal, should be cooler.

    Of course, the surface temperature of Venus is 735.5 K on average. How about that? Why do you suppose that is?

  30. 130
    dcomerf says:

    Does anyone have a link for any interesting speculative natural histories of Venus?
    i.e. presumably the slow rotation and the approx zero magnetic field are linked? If Venus always rotated this slowly would water not have boiled on it’s day side in a pre-high-albedo pre-super-greenhouse point in history?

    Would any decent atmospheric water content combined with no magnetic field lead to the vast bulk of its hydrogen being lost to space and the supergreenhouse arising due to no hydrological cycle (rather than some runaway point when previously stable oceans reached boiling point)?

    What mechanisms could have caused some putative early Venus with a “more normal” rotation period and a magnetic field (and potentially a biosphere) to change into the present very slow rotation no magnetic field state? If the only answer is massive asteriods etc could any biosphere have survived the magnitude of the impact required to make the change anyway?

  31. 131

    dcomerf 130,

    The present theory is that the rotation of Venus occurred through a stochastic process–it was the accidental result of the last big impact.

  32. 132
  33. 133
    kenny says:

    Hey very nice blog!! Man .. Beautiful .. Amazing .. I will bookmark your blog and take the feeds also…

  34. 134
    AlC says:

    The Uncertainty Prayer should be posted prominently on the walls of all policy analysts/decision-makers.

    Few people posting in this thread actually are speaking to the content of the referenced white paper. I found it interesting reading.

    Not being an expert on GCMs, I would be very interested to read the comments of any of the resident gurus (Gavin?) on any of the points in the paper.

    3.1.8 How well do the models reproduce past climate, variability, and extremes?

    The paper also offers options for investments in modeling and data gathering to reduce uncertainty for regional forecasts, but with a range in climate sensitivity of 2 – 4.5 C/ doubling of CO2e, and uncertainty in levels of future emissions, it becomes apparent that the best that can be hoped for is a substantial range…which brings us back to the Uncertainty Prayer.

    And a related prayer that decision-makers see the need for reducing emissions to near-zero ASAP, and find the wisdom to make it happen.

  35. 135
    Ric Merritt says:

    Septic Matthew @ 2 July 2010 at 5:57 PM (#108)

    declines to bet on future sea ice because

    “No. I claim no knowledge about future climate, other than that weather will fluctuate.”

    Nuff said. You original remark that elicited my bet proposal was deliberately misleading. Coming from a long-time participant here, offensively so.

    Other people do quite reasonably claim *some* knowledge about future climate. The usual uncertainties cited do not cancel all knowledge, or if you prefer, predictive power. Your retreat to know-nothingness is a miserable cop-out made inevitable by starting from an untenable position. (Hint: to those of us with time served in academia, “untenable” is way worse than “wrong”.)

  36. 136
    Anonymous Coward says:

    AIC (#134),
    I’m obviously no guru but check the “index” link at the top of the RealClimate home page. The stuff under the heading “sensitivity” is relevant to your query, especially the “plus ca change” entry. I also recommend the “target CO2” and “methane hydrates” entries for additional uncertainties not taken into account in your quote.

  37. 137

    OT–but so gloriously so that I hope the ‘fun factor’ will justify the comment.

    The Solar Impulse team just completed a 24-hour solar-powered flight over the Jura mountains. What a great stunt demoing advancing PV capability this is!

  38. 138
    Martin Smith says:

    There is a heat wave in the eastern US, and everyone is aghast. But the temp anomaly map indicates there is no anomaly there. Is everyone just forgetting that it gets hot on the east coast in the summer, or is the anomaly map wrong?

  39. 139
    Rick Brown says:

    #138 Martin Smith:

    It might help to try this instead (courtesy of Wayne Davidson on the Muir Russell thread)

    You can stop the animation and advance one day at a time through the relevant days — July 4 -8.

  40. 140
    flxible says:

    re 138&139 – I’d say the animated one got it close for my area, very anomalous for early July, can’t get anything done out in the sun!!