Fractal Weather Prediction

Judebert writes "Benoit Mandelbrot, the man responsible for much of the interest in fractals, spoke last month at the American Giophysical Union meeting. He explained how he has been using fractals to find order within complex systems in nature, such as coastlines and weather. (I thought he was dead, but apparently he's just been teaching at Yale.) Earth scientists have taken his fractal work to the point of forecasting the size, location, and windspeed of hurricanes at landfall. Their predictions are being made available to FEMA and other government agencies."

I wonder if there's a real independent assessment

[SIGFPE]

to see how well this technique actually works. Dr. Mandelbrot is, above all, a very good seller of himself.

Re:I wonder if there's a real independent assessme

[Dancin_Santa]

Personally, I'd like to see the results of applying his technique to earthquake prediction. Above all else (hurricanes, tornados, snow storms, volcanic eruptions, etc.) earthquakes have the possibility of doing the most damage to a large area in the shortest amount of time. Being able to predict the location, time, and magnitude of them would be a boon to human existence.

As for myself, I have no need for weather prediction. Cold today, cold tomorrow. The five day forecast: cold cold cold cold cold.

Dancin Santa

does anyone else find this funny?

[delorean]

I thought he was dead, but apparently he's just been teaching at Yale.
To me, that is just too funny.

Prolly' not to any Yale w/men in the audience, though.

Red Spot

[Perdo]

This could be applied to Jupiter's red spot during the upcoming collision with the white spot. That would certainly give insight into Earth's hurricane formation zones and perhaps predict how global warming could aggrevate storm strength.

Dead? Or Worse?

[mfarah]

(I thought he was dead, but apparently he's just been teaching at Yale.)

Of course, he wishes he were dead. }:-> .

Earth scientists have taken his fractal work to the point of forecasting the size, location, and windspeed of hurricanes at landfall.

Maybe Mandelbrot's death rumours started with an alleged claim of him doing field work?

Re:Dead? Or Worse?

[Negadecimal]

Of course, he wishes he were dead. }:-> .

Perhaps he *is* dead. Imagine it: this "Dr." Barton figures he has more credibility if Mandelbrot backs his work. So he stages his own little Weekend at Mandelbrot's.

Anyone ever wonder if Stephen Hawking has just been incredibly well-preserved, and all the while, some junior physicist remotely hijacks his wheelchair and voice synthesizer?

Talk about morbid paranoia ;)

Chaos Weather

[cmpalmer]

Of course, the only problem is that you have to have accurate census and location information for every butterfly on the face of the planet in order for the forecast to be accurate.

Re:Chaos Weather

[alawi]

it depends on how far in the future they are predicting the weather. Does anybody know how long it takes before the minute "butterfly" difference can become a hurricane-sized difference in our happy chaotic system? I expect that it's a long enough time that they don't have to worry about it, but it will limit the technique...

Probably using Fractal Statistics

[Jormundgard]

I'm guessing that this guy is using "fractal statistics" to describe the probability of the atmosphere moving into different states. I don't think that this necessarily implies that "The weather is fractal" or self-similar or whatever. The fractal statement is ususally at best an approximation anyway, and at worst completely untrue :).

But don't trust me too much on this (A good policy when reading slashdot posts!). All I know about fractal statistics is that it uses fractal function as it's distribution. Maybe someone else knows more about fractal statistics?