Weathering Heights : Interview with Edward Lorenz

2023-05-09

Feature Interview with Edward Lorenz, PBS Springboard, ????-??-??

a few links added by Tor G. Syvertsen

In 1960 meteorologist Edward Lorenz (1917-2008) happened upon the discovery of chaos theory and that one discovery has profoundly changed how we look at the prediction of weather.

I recently spoke with Dr. Lorenz from his studio in Boston at Massachusetts Institute of Technology and began by asking him to describe that discovery 40 years ago.

Edward Lorenz: Well, I was working with a mathematical system, a few equations that were supposed to represent the behavior of the weather.

And this was on a small computer and when I ran it a second time, it didn't do the same thing the second time or, as I thought it should have.

Then I found that some of the numbers that I put in the second time were rounded off more than they had been the first time.

And this little bit of round-off was enough to— so that in a simulation of a couple of months the weather was doing something entirely different.

Rebecca Roberts: So when you initially put in 6 decimal places it came out to one outcome and again with three decimal places it was a dramatically different outcome.

Edward Lorenz: That's right, yes. For a while, it was undetectable it was the same as far as we could tell and I could see it was slightly different then entirely different. What had happened I had gone out of the room for a while and came back afterwards and it was already the different the first time I saw it.

Rebecca Roberts: Really? What point in the simulation did it start to veer?

Edward Lorenz: it would have been noticeable by a month or a little bit after that.

Rebecca Roberts: So, this has been called in development the butterfly effect, can you explain what we mean by the butterfly effect?

Edward Lorenz: Well, this name, which I think was due to James Gleick, the author of "Chaos: Making New Science" referred to the fact that a— something as small as the flap of a butterfly's wings might in due time stir up something as large as a tornado. Uh, actually it wouldn't stir it up but it would cause a tornado to appear when it might not have appeared otherwise and could equally well prevent a tornado that would have appeared otherwise.

Rebecca Roberts: What's the significance of chaos theory coming out of meteorology as opposed to another science?

Edward Lorenz: The weather, the atmosphere is a chaotic system. All the indications are meaning it is one that does exhibit chaos one of the first things of that sort where the equations, which mostly govern it or a sim play form were put on a computer. This was in the early days of computers when they were just becoming common that we began to realize that the atmosphere behaved this way.

Rebecca Roberts: Since then computers are obviously enormously more sophisticated capable of hundreds of thousands more calculations.

Does that mean we are now more capable of approximating true weather prediction or does the chaotic nature of it mean we'll never truly predict the weather.

Edward Lorenz: The atmosphere is just as chaotic as it ever was.

We still don't predict it as accurately as I think we are going to be able to in some time but there's going to be a limit to however— how well we will ever predict it say a month ahead because we can't — we can't examine it.

We can't examine all these little butterflies all over the earth to see what they are doing and their influence begins to build up at about that time.

Rebecca Roberts: so, when you first developed this chaotic theory of weather prediction you thought maybe it meant we couldn't ever predict the weather accurately. Have you changed your mind about that?

Edward Lorenz: well, my feeling then was that we couldn't predict it at sufficiently long range. We certainly can predict it pretty well a day or two ahead as we were doing moderately now even then although we do much better a day ahead now than then. I haven't changed the feeling to say we can't predict individual storms a month ahead. I still don't think we can.

Rebecca Roberts: Thank you so much for being here on "Springboard".

Edward Lorenz: You're very welcome. I'm happy to be here.

Edward Lorenz biography

…