The Big Picture
Although immediately mired in politics, I nonetheless felt compelled to note that on Wednesday a 5-year joint effort between the UK and Japan will begin, first to load the two most sophisticated global climate mapping programs in existence onto a supercomputer called the Earth Simulator, and then by next year to combine them into an even more effective program that only this computer (currently) would be able to run.
Because the issue - at its core the global warming issue, which in turn has at its core the question of to what degree modern human industrial endeavors affect global climate - is so contentious some are rushing to wave this project off as unimportant. Indeed, if it were to turn out to have simply been a tool to support global warming theories rather than a scientific endeavor to more precisely understand the points of balance at work, I'd agree. Still, while there is a basis for concern that those involved in the project may already have reputations invested in global warming theories, and so may bring that bias with them to the construction of larger models, in the long term that should become unimportant. The attempt to more thoroughly map ongoing changes and the degree to which individual factors appear to be affecting them is a good direction to go. It's certainly a more scientific one than simply declaring it political and avoiding the questions.
The extent to which it will provide practical information within the 5-year plan of the project is, of course, unknown. Climate is a long term affair, and so systems of this sort will have to be free to explore and project down multiple paths, and then we'll have to allow the time to pass to see which ones are born out by actual events... of course, the extent to which human affairs do or don't change factors over time, those paths will be up for constant revision. It's a complicated affair.
The article seems to sweep aside the fact that these climate models are built on meteorological models -- that is, in the short term they deal with the weather. It occurs to me that keeping that aspect in view should be a key facet of this. Not only would that provide a greater potential for pragmatic spin-off information (more precise global weather tracking and prediction is of direct interest to all of us, but it's also of key interest to governments, major corporations and those who insure them) to help justify this work in the short term, but it also would provide a more human-observable feedback mechanism for determining which of the models is most accurate. Certainly, we're not talking about whether or not it's going to rain on New Jersey on August 5th -- that's still too specific -- but rather predictions about the intensity and primary points of generation for hurricane season. If they keep an eye on the results at that level it almost can't help but provide a more informed model.
As it is, the selling point behind the project seems more vaguely political - in terms of international standing - than anything else. This is reflected most clearly in writer McCarthy's (no quotes are employed to attribute this to the UK's Foreign Secretary, so I'm taking this as the writer's perspective) "...put Britain at the cutting edge of foretelling the future the modern way." Certainly, pure science must be seen as an end unto itself, but it always becomes a little muddled when it steps (as it always must, due to the need for funding) into the political arena.
In the short term, especially with its focus on such a long term item as climate, we will have to view any dramatic pronouncements from this project with a healthy degree of skepticism. Hopefully the need to do good science will not be outmaneuvered by either an existing political agenda and/or someone's desire for a sexy, overly-specific headline of doom to feed to the news services.
In the meantime we can hopefully agree that this is a better use of this level of computational speed and ability than were it turned into the latest MMORPG engine.
Also, isn't something of this scale the grist of SF computers-becoming-self-aware stories such as Colossus: The Forbin Project and the SkyNet of Terminator movie infamy? Okay, okay, so those were primarily military systems to start, given control over said weapons. What would GAIA do? Broadcast scary stories? Rerun old Chiffon margerine commercials? (Hey, apparently Dana Dietrich -- Mother N from those commercials -- is still alive. Well, she's only 76 so I suppose it's not that surprising.) No, wait... we could be terrorized anew by Captain Planet. Brrrr.
You people are so difficult on Mondays.
Comments