“It’ll be okay: Trust me”, redux


Professor Steven Koonin offers up another dollop of vague, specious criticism of climate science in his editorial in The Wall Street Journal. He is credentialed, no doubt authoritative. But compelling arguments for a position should be judged as if the speaker’s identity were unknown and, so, for the most part, I’ll leave it to the reader to find out who Koonin is, or was, by consulting the Wall Street Journal article itself. (Hey, they published it. They deserve a few extra clicks to appease their profit-demanding owners. And, while I respect his career, the editorial he writes is in many ways thoroughly disappointing, not reflecting the deep knowledge no doubt Koonin has. Did the Wall Street Journal get nervous about an earlier, more honest version?) No doubt it is entirely coincidental this op-ed appeared on the weekend the largest climate march to erupt on the planet yet. (Professor Koonin could have published it at any time. But nay.) The actual argument is the latest of a long series of reasons for why climate science should not convert to climate policy.

Let’s see what’s been admitted in this rendition of Tom Petty’s “Yer So Bad” (in reference to climate science).

Early on Koonin admits key points, often opposed by so-called “climate deniers” of the past (*):

  • Climates can change, and they can change very rapidly, sometimes as quickly as a decade.
  • Climates change because there are causes. Koonin glosses “causes” by using the term “influence”. That might reveal his prejudice, but it grants the case. Koonin grants that the “impact today” is comparable to natural climate variability. Koonin leaves completely open what the impact might be in future years. (See the discussion of “lags” below.)
  • The “influence” of carbon dioxide emissions will continue “for several centuries”. I fault Koonin seriously here, because there is plenty of evidence that some of the human generated carbon dioxide will remain in atmosphere for millenia, and that science is very basic.

Koonin then proceeds to make a number of claims:

  1. The question of how climate will change over the next century is unsettled science.
  2. “[C]hoices about energy and infrastructure” should be based upon “answers” to that question.
  3. Science does not understand the oceans well enough in order to make such projections.
  4. Climate feedbacks, notably those concerning water vapor feedback, “are uncertain”, and cannot be predicted from ab initio physics, but must be based upon “precise, detailed observations that are, in many cases, not yet available”.
  5. Climate models used to project what climate might be for the next century are imperfect and sloppy. (Note, nothing is said about their ability to project beyond one century.)
  6. The IPCC Summary for Policymakers sweeps under the rug many of these nuances.
  7. Climate science represents “heroic research” which costs “billions of dollars”.

With these claims, Koonin remains in the realm of the sensible and the accurate. Okay, that “billions of dollars” is bait for the Tea Party. Do readers know most scientific participation in the IPCC is on the scientists own dime, including travel costs? And he could suggest that if the climate science is correct or understates the problem, the trillions of dollars it will cost to try to fix the problem. And the “unsettled science” of the next century is not bottom line “unsettled”, it’s poker. But, then, he launches off into Climate Zombieland, claiming:

  1. Computer modeling of complex systems is as much an art as a science. Partly so, but that’s part of science. The same might be claimed of medicine. But most science and engineering is “as much an art as a science”. Does that make buyers of BP gasoline think it’s gasoline is less reliable than Exxon’s? No, of course not. That’s because the quality control needed to produce it, whatever the art and science mix involved, produces a repeatable usable product. The “art” here is not Salvador Dali. The “art” here means “disciplined practice”. Koonin is equivocating the two, and thereby misleading.
  2. One important feedback, which is thought to approximately double the direct heating effect of carbon dioxide, involves water vapor, clouds and temperature. This is patently false, or at least badly misleading. Water vapor can double the direct heating effect of carbon dioxide, but it also removes heat as well. Koonin is telling the reader one side of the story.
  3. [F]eedbacks are uncertain. All of science and engineering is uncertain. But one way of defining science and engineering is that field of investigation where quantitative knowledge is preeminent, and quantifying uncertainty in that knowledge is a basic requirement. This quantitative knowledge is not the fiction of most accounting practice, which parades itself as solid knowledge, but must necessarily be uncertain and is subject to a myriad of assumptions and measurement inaccuracies. In science and engineering, the practice (or “art”) of the field demands such uncertainties be quantified and stated. The Summary for Policymakers leaves specific risk numbers out, trying to replace them by a qualitative and coded schedule of words. That’s by choice of the countries who have final editorial say over the Summary’s contents. If a reader wants to see what the scientists actually say, they can consult the Physical Science Basis, or the section on Impacts, Adaptation, and Vulnerability, also by scientists, or the section on Mitigation of Climate Change. By the rules of the IPCC countries can delete phrases and sentences in the Summary on Policymakers, but they can’t touch the full scientific reports. In fact, much of the “denier arbitrage” that’s played out in the media has people quoting from the Summary sentences which have been redacted or edited by countries.
  4. The models differ in their descriptions of the past century’s global average surface temperature by more than three times the entire warming recorded during that time. Such mismatches are also present in many other basic climate factors, including rainfall, which is fundamental to the atmosphere’s energy balance. Yeah, that’s why an ensemble of models is being used, because of the strong statistical notion that an ensemble is going to have higher predictive value than any single model by itself. It’s the same reason why systems like the Iowa Electronic Markets are seen as having stronger predictive power than, say, election polls.

Koonin goes on, and he gets increasingly misleading. He faults climate models for failing to predict the “warming hiatus” (see here and here), yet says nothing about models which include oceanic effects and succeed in doing so. He faults climate models for failing to predict increase in Antarctic sea ice claiming they “roughly describe” the loss of Arctic ice. That’s grossly misleading and either is cynical or demonstrates ignores. Climate models totally missed the rate of Arctic sea ice loss. Koonin seems to not want to point that out, because it shows that the imperfections in climate models could just as well fall on the side of reality turning out far worse than they predict as it could fall on being better than they predict. And he repeats a denier claim regarding Antarctic sea ice which he should know far better about: Increasing Antarctic sea ice is because the Antarctic cap is dissolving, not because it’s getting colder. It’s comparing apples and oranges. There’s no continent at the North Pole. There is a continent at the South. He’s exploiting ignorance in people who don’t know this distinction. Given his claim to knowledge, that’s just unethical. He also completely skips the fact that we already know, from climate history, that over the next several centuries sea level will rise 65 feet, no matter what we do. All we can do is to help prevent it from happening sooner, and keep it from getting worse.

Being such an expert in computational modelling, Koonin could instruct the audience in the importance of lags in physical systems. That is, if a force or cause is imposed right now, it is not at all the case, especially for extended, complex systems, that the effect will be seen right now, or soon. And, if he were true to his expertise, he might suggest that there is a low but not insignificant probability that the linear models upon which these projections are based could be missing strong nonlinear couplings which might produce massive effects much sooner than projected, in the same manner which they missed the collapse of Arctic sea ice. While Koonin may not be speaking primarily to a science audience in his op-ed, he could recommend concrete improvements like those Palmer did in 1999.

The most outrageous thing about Koonin’s op-ed, for me, is that despite his harping on the ignorance and the need to improve, no where does he recommend a massive scientific campaign to improve our state of the oceans and the climate. Funding for these has been severely cut, and the resources and science for doing them is suffering. Instead, he proposes another layer of review.

Koonin could have done everyone a great service to clarify and unify. But he ended up being “another pretty face, with credentials” arguing “Trust me: It’ll be okay.” We’ve heard that before.

In all, it is a disappointing and bad bit of rhetoric.

But I wonder: Will anti-“climate alarmists” will sleep with a copy under their pillows after they see the coverage of tomorrow’s People’s Climate March?

Update, 10th February 2015

There was an op-ed about Koonin’s perspective by Ray Pierrehumbert at Slate.


(*) Read my next book, A Natural History of Climate Denial, published by Nonesuch Publishers, in the Spring.


Thanks to Professor John Carlos Baez of UC Irvine and The Azimuth Project for indicating a phrasing and spelling problem.


About hypergeometric

See http://www.linkedin.com/in/deepdevelopment/ and http://667-per-cm.net
This entry was posted in art, Boston Ethical Society, carbon dioxide, carbon dioxide capture, Carbon Tax, citizenship, climate, climate education, conservation, ecology, economics, education, energy, engineering, environment, forecasting, geophysics, mathematics, maths, meteorology, oceanography, physics, politics, rationality, reasonableness, science. Bookmark the permalink.

8 Responses to “It’ll be okay: Trust me”, redux

  1. Pingback: Angry Bear » Guest post: “It’ll be okay: Trust me”, redux

  2. m.jed says:

    Is there any specific piece of scientific evidence that could be found, discovered, or released in the next 1-5 years that would be inconsistent with the AGW thesis?

  3. Sure, you could find evidence, but it would be properly doubted. The idea that a single measurement can refute a major theory is based upon a big misunderstanding of how science operates. Scientific theories or hypotheses are not legal arguments or debating arguments, logically based, so that knocking out one measurement causes the whole thing to fall as a stacked deck of cards. It’s a carefully constructed and mutually supporting network of theoretical ab initio calculations, predictions derived from those calculations, empirical measurements, and uncertainty statements for each of these. There is a great deal more than climate which is based upon this science. It’s used in engineering to build spacecraft and to manufacture semiconductors. If the basic science were wrong, these devices could not work. That they do supports the science.

    There’s a formal way of looking at this which is quite consistent with the theme of this blog. That is, to determine the (posterior) probability that the theory (or hypothesis) is true given a datum one takes priors on the parameters of the hypothesis and the Likelihood of the datum given the hypothesis and theory, takes their product, and then divides by a normalization constant. If a theory (or hypothesis) is well supported by lots of data, a single datum is not going to nudge the needle on the plausibility of the theory (or hypothesis). This is as it should be.

    Contrary to the legal kind of argument, standard science demands of critics of well accepted theories (or hypotheses) that they need to come up with an alternative construct, an alternative hypothesis which explains all the prevailing data better than the existing theory (or hypothesis). That’s what relativity did. That’s what quantum theory did.

    Taking Kuhn’s Structure of Scientific Revolutions, most scientific effort goes into normal science, which concerns the verification of existing scientific hypotheses. Revolutions happen only when existing explanations (like the “ultraviolet catastrophe” in the case of blackbody radiation) have a big problem, and a new explanation comes along which explains all the data (as I already said) as well as data beyond the existing.

    Do you now understand?

  4. Eli Rabett says:

    Koonin is, to put it bluntly, opinionated and ignorant about climate. His performance at the APS subcommittee review working on a new public statement was asinine and all of the folks brought in to educate the physicists (all from the nuclear complex) had to explain simple things to him and the rest of the committee.

    His performance is what Myanna Lahsen calls the arrogance of physicists as summarized to her
    http://sciencepolicy.colorado.edu/admin/publication_files/resource-2590-2008.05.pdf

    this is a problem with physicists: they think they know everything, because they’re smart. What they don’t understand is that yes, it is true, actually meteorology is a branch of physics. And so you take a physicist, like me, and you can sit him down, and in 2 or 3 years, they could learn meteorology. But physicists confuse being smart and having the ability to learn everything with actually knowing stuff!

  5. John Baez says:

    Great post – I hope the Wall Street Journal publishes it. 🙂

    One typo: “Being such an expert in computationally modeling” – should be “computational modelling”.

  6. Ray Pierrehumbert, Professor of Geophysics, University of Chicago, weighs in on Koonin’s op-ed at Slate.

    Ray’s book site is Principles of Planetary Climate. Notable are his Python scripts and datasets.

    To give you an idea of how enlightening Professor Pierrehumbert’s teaching is, I own two copies of his textbook, one at home, and one at work, so, if I think of a question, I needn’t wait ’til I get to the other place to answer it. I also work some of his problem sets during lunch breaks.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s