Monday 9 October 2006

Proper science.

Back in June, I expressed a certain amount of cynicism towards the results claimed by scientists who rely excessively on computer models. Despite my saying up front that I wasn't especially picking on climatologists, there was still a bit of an indignant kerfuffle in the comments courtesy of a reader who believed that any even tangential criticism of the humans-are-boiling-the-planet-to-death theory — even just a criticism of the methods that some of its exponents use to reach some of their conclusions — is an ignorant attack on the science of climatology and, indeed, on the whole of science itself.

So it's nice to see this example of climatologists using some seriously impressive science to get some proper results:

A team at the Danish National Space Center has discovered how cosmic rays from exploding stars can help to make clouds in the atmosphere. The results support the theory that cosmic rays influence Earth's climate.


The experiment called SKY (Danish for "cloud") took place in a large reaction chamber which contained a mixture of gases at realistic concentrations to imitate the chemistry of the lower atmosphere.

Ultraviolet lamps mimicked the action of the Sun's rays. During experimental runs, instruments traced the chemical action of the penetrating cosmic rays in the reaction chamber.

The data revealed that electrons released by cosmic rays act as catalysts, which significantly accelerate the formation of stable, ultra-small clusters of sulphuric acid and water molecules which are building blocks for the cloud condensation nuclei. A vast numbers of such microscopic droplets appeared, floating in the air in the reaction chamber.

"We were amazed by the speed and efficiency with which the electrons do their work of creating the building blocks for the cloud condensation nuclei," says team leader Henrik Svensmark, who is Director of the Center for Sun-Climate Research within the Danish National Space Center. "This is a completely new result within climate science."

Now, this result is going to get jumped all over by some very enthusiastic people who think that global warming is complete bollocks but know very little about science. I'm not one of them. This is a very new result, and it would therefore be foolish to go drawing too many conclusions from it. For all we know, for various reasons that no-one has even thought of investigating yet, this makes it even more likely that man is catastrophically heating the Earth and needs drastic action to save future generations. (For the record, my main disagreement with the global-warming crowd is over what that drastic action, if it prove necessary, should be.)

No, I'm drawing attention to this news for two reasons.

Firstly, we now know for a fact that every single climate model ever developed, from the basic ones to the very best of the best, omitted a major and significant piece of information about what shapes our climate. It was omitted simply because no-one knew it. It would be very surprising indeed if more scientific discoveries about what affects our climate aren't made in the next century. That doesn't mean that there is no point in using computer models in science — they are very useful indeed. But it does remind us of how rash it is to make a prediction of the future based on a model of a system that we do not fully understand. This isn't a criticism just of the theory of catastrophic man-made global warming, either: all those climate models that showed that the Earth was cooling down, that it was staying the same temperature, that it was warming up but not because of man... all those models are now every bit as obsolete as the ones which blamed America for plunging us all into a fiery death, or something.

Secondly, this is a great example of a proper climatology experiment. What's happened here is that a group of scientists have discovered what happens when the gases of the lower atmosphere interact with ultraviolet light and electrons, and they can be confident that they're right because the way they did it was to take the gases of the lower atmosphere, some ultraviolet light, some electrons, and mix them all up and watch to see what happened. And here's a thing: we know enough about electrons to model them accurately, we know enough about the various gases in the atmosphere to model them accurately, and we know enough about light to model it accurately, yet no-one got this result from a mathematical model. In fact, if this result had come from a model, it would still be regarded, quite rightly, as a hypothesis, albeit one with a little more evidence in its favour; it is only once the result has actually been replicated in reality — as it has — that it attains the exalted status of fact.

Over the years, climatologists have often used similarly robust methods to develop the theories behind global warming — how do you think we know that carbon dioxide contributes to the Greenhouse Effect? — and those theories are generally solid and sound. At other times, climatologists have used computer models to make predictions about what will happen a hundred or more years in the future. There is a reason why some of us place a lot more trust in the former and cynicism in the latter, and that reason, thank you very much, is not ignorance of climatology, ignorance of computer modelling, or ignorance of science itself. On the contrary, it is scientific ignorance that leads the general public to give both types of result the same weight.

No comments: