The Butterfly Effect: For Real?

In Jurassic Park, when Jeff Goldblum blabbered on about Chaos Theory and butterflies causing hurricanes, I thought he was full of it. Turns out that butterflies do affect the weather, at least according to the EPA’s EnergyStar PSAs.

I heard this little gem this morning on the radio:

Think I’m making this up? Download the mp3 of the PSA here. According to the EPA’s website, “451 stations are airing the ‘Butterfly’ and ‘Breathe’ radio spots in English and Spanish in 263 cities and 49 states.”

So is it true? Are butterflies really a meanace to us all? Or is the EPA actively spreading stupidity?

This sounds like a question for Cecil.

I think Chaos Theory (in this instance we’re talking about a subset called Complexity) is being a little misunderstood here. There is no causal relationship: If butterflies disappeared, no more or fewer hurricanes would result.

The point is that if you worked backwards from the hurricane, trying to quantify the position and velocity of every water molecule in order to “explain” the hurricane, then you would find that a tiny change in some of those initial variables would have macroscopic effects in how the weather developed.

Using this idea to suggest that saving energy might produce fewer hurricanes is a little misleading, however since we do not know how much effect adding CO2 to a complex system balanced at an apparently delicate equilibrium will have on weather patterns in future, perhaps the parallel is justified.

Homebrew, this is an often misunderstood concept. It’s not like if we had no butterflies there would be no hurricanes, or everytime a butterfly flaps it’s wings it has any real effect on weather patterns.
Weather is a chaotic system, which means that small, even minute change in the intial state of the system can, over time, lead to huge changes.

So yes it is possible that a hurricane could happen that wouldn’t of happened if every other factor played out exactly the same because a butterfly flapped it’s wings or didn’t flap it’s wings.

The butterfly effect is discussed at length in this book on chaos theory.

Also see here

If I remember correctly from the book…even if you were able to place weather instruments (temperature, pressure, humidity etc) 1 meter apart in every direction of the atmosphere…one would not be able to make really long range forecasts because of the microchanges that occur within that 1 meter space.

I think this discovery describes the concept well.

Like others have said, the butterfly effect is not meant to be taken literally. It’s more an illustration of one of the basic ideas of chaos theory.

It is true in a truely chaotically system. However, is the weather actually truely chaotic?. Eg. perhaps the energy of a butterfly gets damped out?

For anyone who’s interested, I gave an explanation of what the butterfly effect represents in this thread.

Well, this is a good question. I believe (although I can’t say I know for sure) that even in a chaotic system, there may be some small perturbations that can get damped out.

So, do we really need to know where all the butterflies are to predict the weather (at some fairly long range in time)? I haven’t heard a good answer to this.

By the way, for moderately long range weather prediction (on the 6-10 and 8-14 day scales), the Climate Prediction Center (CPC), which is part of NOAA, is actually running “ensembles” of simulations using their forecase models. The ensemble members differ by perturbations in the initial conditions. This allows them to see how much they can trust the predictions. For example, sometimes the different ensemble members still are fairly tightly clustered around the same solution after 14 days but other times they are very spread out or cluster around a few different solutions. Thus, they will not only issue predictions on a 6-10 day and 8-14 day scale but will also use this and other information to discuss the confidence level for their predictions.
See and look at the definition for “GFS ensemble”.