The Island of Doubt

Science as a Contact Sport:
Inside the battle to save Earth’s climate

by Stephen Schneider
National Geographic, 295 pages

Not even Stephen Schneider could have anticipated how timely his new book, Science as a Contact Sport: Inside the battle to save Earth’s climate, would be. The histrionics generated by the theft and publication of the UEA emails suggests climatology is much rougher than even Australian rules football.

Schneider was one of the first climatologists to understand the need to communicate what his research was showing with the general public. He appears in documentaries back in the 1980s, and is second only to Jim Hansen when it comes to battle scars. I dout there’s anyone better positioned to tell the inside story of how science is hammered out.

I read this book a few weeks ago, before the purloined emails found their way onto a Russian server, which helps explains why I used words like “banal” to describe the correspondence that so many climate change pseudoskeptics find so shocking.

It also helped that I have spent several years actually working along actual scientists. So the notion that they can sometimes be arrogant, hypersensitive, snarky and impatient with their critics did not exactly comes as a surprise. Schneider’s book, however, provides some rare insight into the challenges that face climatologists who are being asked to do much more than just scientific research.

The IPCC process, for example, demands much of the volunteer army that produces the reports we use as the definitive, albeit politically watered down, word on the state of climate science. Here’s what Schneider has to say about his introduction to that particular realm:

The way some IPCC detractors tell it, the IPCC is a loose cannon of radicals, which is nonsense — it is much more a debating club dominated by conservative empirical scientists. It took several assessments just to get them to mention surprises, let alone formal subjective probabilistic treatment of uncertainties.

Schneider revisits the war between empiricists, who tend to scoff at the modelers and whom they accuse of violating the core tenets of science. The modelers, on the other hand, can’t understand what’s wrong with using the only tool at their disposal to predict the future, which is, after all, what anticipating climate change is all about.

This sort of debate continues today. For example, a pair of papers released last month within a week of each other came to contradictory conclusions. One, in Nature Geoscience, says the airborne fraction of carbon dioxide is increasing, meaning that carbon sinks are loosing their ability to soak up our emissions. Another, in Geophysical Review Letters, says the airborne fraction hasn’t changed at all. The difference is the GRL paper, by Wolfgang Knorr, is based on empirical data while the Nature Geoscience paper, by Corinne Le Quéré, uses models to come up with the estimate.

It’s this kind of thing that consumes climatologists. When, that is, they’re not fulminating against the pseudoskeptics who think no one has anything better to do than respond to freedom of information requests for data that’s already freely available. Science as a Contact Sport covers both the internal bickering and the circle-the-wagons mentality that many believe is at least partly responsible for the public relations mess in which the entire field is now mired.

If climatologists are only human, so are journalists, however, and Schneider doesn’t restrict himself to his own profession. He devotes more than a few pages to the consequences of giving a nuanced quote to a journalist. He told Discover magazine:

So we have to offer up scary scenarios, make simplified dramatic statements, and make little mention of any doubts we might have. This “double ethical bind” we frequently find ourselves in cannot be solved by any formula. Each of us has to decide what the right balance is between being effective and being honest.

That part of the quote gets a lot of mileage in the denialosphere. But the full quote ends with a line that is almost always left out.

I hope that means being both.

It’s no wonder that some scientists are leery of talking with reporters.

Fortunately, Schneider is wise enough to know someone has to keep talking. And he does. His story is his own; no doubt other participants in the conferences he attended would remember things differently, but that’s beside the point. What’s important is not who’s right, but how science is done. And anyone who takes the time to follow Schneider down memory lane will have a hard time believing that climatologists are part of some global conspiracy to defraud the public.

Unless of course, you think Schneider is in on the conspiracy.

Ordinarily, I’d wrap up a book review with something from the book. But Peter Watts, a marine biologist who now writes science fiction, recently offered a take on the email controversy that helps explain why books like Schneider’s have never been needed more than today. Schneider doesn’t come right out and say it, but I think he would agree that:

Science doesn’t work despite scientists being asses. Science works, to at least some extent, because scientists are asses. Bickering and backstabbing are essential elements of the process. Haven’t any of these guys ever heard of “peer review”?

If only this book had come out a few months ago. It wouldn’t have deterred too many of those who are already convinced all climatologists are evil communists. But maybe a few critics would have understood where the climatologists are coming from and not been so quick to judge.


  1. #1 bob koepp
    December 4, 2009

    Obviously, it’s desirable to be both honest and effective. When there’s a tradeoff however, I would hope that scientists (and journalists) would plump for honesty. Scientists (and journalists) who are conflicted on this point should probably consider alternative occupations — perhaps marketing would be a better fit.

  2. #2 Lyle
    December 4, 2009

    The behavior seen is entirely consistent with the structure of scientific revolutions in that climate science has a strong paradigm right now. Perhaps the whole episode could be used to teach the media the about Kuhn’s work. Its been a long time since high school, but I sort of doubt that the Structure of Scientific Revolutions is taught much in high school, which means that the public does not really know how science is done. The public views it as a bloodless no emotions as some one put it Vulcanian (as in Spock) profession. But Scientists are not Vulcans but humans and have emotions. In some cases a scientific revolution takes a generation as new folks evaluate the evidence and come to a new conclusion. What I would use as material for examples is the earth sciences in particular the Spokane Floods, and the plate tectonics revolutions.

  3. #3 Pierce R. Butler
    December 4, 2009

    About 25 years ago, I read a piece by Schneider (in Scientific American, IIRC), saying that meteorologists had developed software that could deliver an accurate 24-hour forecast – but that it took the supercomputers of the time 48 hours to crunch the numbers.

    If he mentioned Moore’s Law, I don’t remember it. By my rough calculations of same, a comparable modern computer should do the job in under three seconds, yet somehow the accuracy hasn’t trickled down to local forecasts yet. Does Science as a Contact Sport address this anomaly?

    Lyle @ # 2: Kuhn’s own paradigm (judging from commentary by working scientists at sciblogs & elsewhere) is not exactly the last word on “how science is done”. The real thing seems to be a bit messier…

  4. #4 Ambitwistor
    December 4, 2009


    Supercomputers today are of course vastly faster than they were 25 years ago. I am sure that “accurate” by Schneider’s 1985 standards meant something entirely different than “accurate” does today. As forecasts improve, our demands for accuracy become more rigorous.

  5. #5 dhogaza
    December 4, 2009

    If he mentioned Moore’s Law, I don’t remember it. By my rough calculations of same, a comparable modern computer should do the job in under three seconds, yet somehow the accuracy hasn’t trickled down to local forecasts yet.

    Really? I live in the US Pacific Northwest, which is notorious for being a place where weather is extremely difficult to forecast. Actually, I live in Portland, OR where wet pacifc fronts often battle artic highs leaking westward through the Columbia Gorge. Depending on timing and relative strength, this can lead to 33 degree rainy or snow mixed with rain days, or 30 degree snow. Or if it leads to an inversion, an ice storm (‘silver thaw” as we say, since they typically come after periods of sunny but freezing days), or just cold rain.

    Modern forecasts here are *extremely* accurate over 24 hours, and not bad out to about 5 days.

  6. #6 llewelly
    December 5, 2009

    Short term numerical weather forecasting is an initial-value problem, and thus extremely sensitive to the quality of the observations used to initialize it, the resolution at which the simulation is performed, and the completeness of the physics. Doubling the processing power available only enables increases in resolution. Numerical weather forecasting simulates physics in 3D. In 3D, halving each dimension of a grid box reduces the volume of a grid-box by a factor of 8 (since 2^3 = 8). Additionally, it requires the length of the timesteps used in the simulation to be halved. Thus, 16 times as many grid-boxes must be processed per unit of time. On top of all that – halving each dimension of the grid-box size only extends forecasts about 24 hours into the future – and even then, only if you already have sufficiently good observations and sufficiently good physics.

    P.S. preview works fine on most of the other scienceblogs – why not here?

    (Long term weather forecasting, like climate modeling, is a boundary-value problem, and faces somewhat different issues.)