Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
On the one hand, true. On the other hand, bollocks.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



affect the weather

Use computation and the chaotic nature of weather to improve weather
  [vote for,

Since our ability to model and predict the weather is improving (and often quite good already), and since computation is continuing to get cheaper, this suggests the following way of taking advantage.

There are things which man could do which might affect the weather. The (not very successful) attempts to do iodide seeding of clouds to cause rain in the 50s are an example. But one can imagine "painting" the tops of clouds to change their reflectivity, or flying planes to have a modest effect on air movement, or other things like that.

So, suppose we run many copies of a weather model in parallel, each modified so that its initial state is the "current" state modified with the effect of some putative change caused by man. If we now run those models, it is possible that one of them will show a state 24-48 hours later which is in some sense "more desirable" than the other models. I use desirable here to mean, say, less rain in an already flooded region, or more rain in a drought area, or something like that.

If the computation is fast enough and reliable enough, one can imagine now performing the action which corresponds to the more desirable outcome. In fact, to account for the time to run the model, you'd want to start it and run it forward for some time, then inject the man-made input, and then run it some more, so that by the time you knew which effect you wanted, the state at which the weather modification was applied was still slightly in the future.

Probably, most modifications would have little or no effect, and it would take some time and experience to learn what things we could do which had a lot of "leverage". Additionally, the weather models would have to be fairly reliable and numerically stable.

capek, Sep 08 2002


       Hm, yeah, we can just get all the people in China to release butterflies at the same time, and we'll have clear blue skies for days.
DrCurry, Sep 08 2002

       Despite your passionate musings, this is simply not possible.
globaltourniquet, Sep 08 2002

       the way meteorologists model the weather is to run one simulation based on the current situation. Then they run hundreds more, each time changing one variable or another by an incredibly small amount. The tendency is that most of them will appear to do the same thing for, say, twenty-four hours, after which point the predictions will suddenly start to point to all sorts of different outcomes. The length of the coherence (my word, not theirs) between the predictions imply their accuracy.   

       Now, given that the methods you propose to change the weather are unlikely to change conditions any more than the deliberate changes meteorologists make to successions of weather models, you may see that they are unlikely to change the weather within the first day/three days (or however long this coherence factor may be), and the way in which they change the weather after that point will be fairly unpredictable. Ain't chaos a bitch?
yamahito, Sep 08 2002

       Actually, given the central importance of the Koch Curve to the subject, I believe chaos is a bastard.
General Washington, Sep 08 2002


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle