Unknown knowns and known unknowables

Posted February 6th, 2005 by Sylvia S Tognetti and filed in Ignorance of Ignorance

(that are unbeknownst to Bush and that are obfuscated by climate skeptics)

Data from Antarctic ice cores now demonstrate that changes in climate are beginning to go beyond the range of variability known to have occurred over the past 400,000 years. The implications of this, as stated in a book (executive summary available here (pdf) published in January 2004 by the International Geosphere-Biosphere Program, is that the “uncertainty, magnitude and speed of change is unprecedented.” Which brings us into “terra incognita” or Post-Normal Times. So kudos to the Bush administration for recognizing and pointing out that changes in climate are uncertain. Under a business as usual scenario, which assumes the continuation of current trends, the climate will only become more uncertain.


This should not be news to anyone. However, given the way uncertainty is often used – to avoid making decisions and policy commitments, this point tends to get obfuscated by the soundbites that make up the discourse of policy debates. It is also given lower priority than sticking to a good and familiar story line. On the other hand, if science could provide certainty, decisions would just be a problem like that of rocket science. With enough research, computer models would tell us the best course of action, and the losers could all be compensated. And if you believe that is even possible, you probably also believe there was a decisive victory for Bush in the recent US election.

Missing from all of the discussion of uncertainty regarding the causes and rates of climate change is an acknowledgement of the numerous different kinds of uncertainty. As is explained at RealClimate, it is important to distinguish uncertainty in climate models from uncertainty about future emission scenarios used in the models. The latter ultimately depend on human behavior and choices made regarding development patterns and technological change, not to mention unforeseeable events such as volcanic eruptions and economic collapse – as occurred in the former Soviet Union. There is also more confidence in projections of global average temperatures, than about more specific changes and their impacts at regional scales, which requires the consideration of site-specific randomly timed events such as large storms and hurricanes, and the vulnerability of actual places. For example, small changes in water levels in the Great Lakes, that might have been inconsequential in another era, could have very significant consequences, because of their direct importance to the shipping industry, which is in turn essential to the economy of the upper-mid-western United States.

A less discussed source of uncertainty, is whether all obtainable scientific information would actually make any difference in policy decisions and actual practices, and ultimately, with their consequences. This goes well beyond issues of whether or not estimates fall inside or outside acceptable margins of error. Use of science to support policy decisions implies also the need to understand the often rapidly changing conditions to which those estimates presumably apply, and therefore, to make judgments not only about the technical quality of information, but also about whether it is even relevant to the questions being asked, and whether the right questions are being asked. Whether outcomes are achieved also raises issues of trust and cooperation, and whether promises are kept. These questions cannot even begin to be answered unless there is some semblance of accountability, as well as agreement about the scope of the problem itself, and visions of the future.

Although not science in the technical sense, making judgments about quality of information is very much part of scientific practice. Inherent in models are assumptions and informed judgments of individual investigators, which must also gain acceptance in the larger scientific community. This is an ongoing and contested process normally carried out through peer review in which there is no final arbiter – at least until science enters the policy arena, when high stakes decisions have to be made, often in response to the unanticipated consequences of previous ones, and to entirely new, or what we will call Post-Normal kinds of problems.

This has implications for the practice of science itself, which then becomes a process of managing uncertainty, rather than one of defining problems so narrowly so as to simply exclude it. In this Post-Normal approach to Science, originally proposed by Silvio O. Funtowicz and Jerry Ravetz, reframing of the problem to consider the broader context, and thereby to include uncertainty, decision-making becomes a process of inquiry and learning, in which the diversity of perspective is to be valued, because it can lead to greater understanding as well as open up more options and possibilities for action. The participation of all affected, or whose behavior can affect the outcome, therefore becomes essential, not simply as a public relations exercise, or for purposes of communicating science and creating “ownership” of the problem, but because it is necessary to understand the problem, correct errors, and develop a broader range of response options.

To return to the election analogy – the outcome of a vote count depends not only on the counting of votes and the technical accuracy of a machine, but also on the rules for counting them, on the integrity of those who operate and monitor the machine, and on whether there are checks and balances needed to detect errors and hold them accountable. Whether people participate and vote may also depend on their confidence in the process and whether they believe it will make any difference in the outcome. Given the impact that presumably elected officials can have on the climate policy, whether votes are even counted is also a source of climate uncertainty. So, at the risk of sounding like a worn out campaign slogan, if Bush and the so-called climate skeptics wish to debate climate uncertainty, BRING IT ON!!!!

The real kudos go to RealClimate for attempting to present an open and accessible discussion of what is and isn’t understood about the science of global warming, and some much needed context for understanding scientific debates.

2 Responses to “Unknown knowns and known unknowables”

  1. Bruce Frykman says:

    >The real kudos go to RealClimate for attempting to present an open and accessible discussion of what is and isn’t understood about the science of global warming, and some much needed context for understanding scientific debates.
    RealClimate.org is certainly accessible, but it is not open. The site is edited and moderated by paid partisans for the catachism of climate castrophy.
    I have tested the thoery that popular critics of this catachism (Sen Inhofe, George Will, Micheal Chricton) receive promenent disparagement for their “unlearned” criticisms while my references to critics from within its IPCCs sanctified halls (Dr Chris Landsea et al) have been discarded by RealClimate.org.
    This site lacks credibiliy

  2. Somehow I lost (perhaps accidentally deleted?) a comment on this post from William Connolley, which was a reponse to a question I asked him on “ask Stoat” (on his old blog, at: http://mustelid.blogspot.com/2005/08/ask-stoat.html. So I retrieved it from there and include here my question and his reply:
    Sylvia said…
    There was an IGBP book, (of which I only read the downloadable executive summary(pdf) because it is expensive) that I’m not sure if I cited carelessly or not. It says the “earth system has moved well outside the range of natural variability exhibited over the last half million years at least.” Although it points out that global change is “more than climate” the case appears to be based on the Vostok ice core data regarding GHG concentrations, and temperatures “inferred” from that. Given all of the fuss and hysteria around the so-called hockey stick, which just examines a much shorter and more recent period, and since I have cited this on my blog – I’m wondering if I need to add some qualifications to statements I have made that we are outside the known range of variability known to have occurred for longer than the period in which humans have existed. So I would appreciate any comments you might have on what evidence there is for climate variation further back than could be reconstructed for the hockey stick. And thanks for your blog!
    11:20 PM
    Belette said…
    Sylvia – well, “earth system has moved well outside the range of natural variability exhibited over the last half million years at least.” I find that a bit surprising. They seem to be basing that mostly on CO2 levels – for which its certainly true. But I think you would have a hard time justifying it for temperature levels: the Eemian (previous interglacial) is generally regarded as being somewhat warmer than we are now. Even for the holocene there were earlier warm periods: see http://en.wikipedia.org/wiki/Image:Holocene_Temperature_Variations.png. These are based on proxies of course.
    Its true the current warming is rapid, though. But I have trouble understanding the all-coral-is-going-to-die arguement, for example: why didn’t it before?
    I found your blog post: http://www.postnormaltimes.net/blog/archives/2005/02/unknown_knowns_1.html. I’ll paste this in there.
    10:06 AM
    Sylvia said…
    Thanks for that – just when you think you’ve finally found a good soundbite – boom! complexity bites again! I’ll eventually work some comments on this into a post. Lesson learned: Soundbites are always dangerous. That is why we need blogs. I may be back here after I check in with a few coral reef wonks…

Leave a Reply