Chaos

How much do you trust the weather forecast?

In the 1950s, scientists were highly optimistic about the possibilities of predicting – even manipulating – the weather. This hope lay in new computer technology.

Of course, they knew that it was hard to get perfect measurements on something as complicated as the weather. But they thought that with good enough data and a lot of computer power, it would be possible to calculate the weather for months ahead – at least roughly.

They’d no idea how fragile, unstable, and chaotic physical systems like the Earth’s weather really are. It took a mathematically-minded meteorologist to demonstrate this.

Here’s the key message: Meteorologist Edward Lorenz became the intellectual father of chaos theory after discovering the unpredictability of weather.

In 1960, Edward Lorenz began running a weather simulation on his brand new computer. He wanted to study how weather patterns change over time. And he stumbled on something deeply unsettling.

Lorenz’s weather simulation was pretty simple – it didn’t even have clouds. Conditions like temperature and airstream were represented by numbers. To study how they behaved over time, Lorenz would pick one of those variables and print out a graph that plotted its fluctuations.

One day in 1961, he wanted to rerun a simulation from the day before. But he decided to start in the middle of the simulation, typing in the numbers from the previous printout by hand.

At the beginning, the second simulation behaved just like the first. But then, the variables’ behavior started deviating. As simulated time went on, they got more and more out of sync. Finally, the motion of the second graph looked totally different from the first.

What caused this massive incongruity? Lorenz had typed in the numbers from the previous simulation only up to the third decimal point. For airstream, for instance, he’d typed in .506. But the computer’s calculations actually ran up to the sixth decimal point: .506127. Somehow, this tiny difference was enough to throw the weather prediction completely off the previous track.

Lorenz was shocked. Like other scientists at the time, he believed that small fluctuations didn’t have big effects on large-scale systems like the weather. Instead, his mistake revealed how unstable, unpredictable, and chaotic these systems really could be.

Lorenz dubbed it the butterfly effect. This means systems like our weather are so sensitive to small disturbances that a butterfly flapping its wings in Beijing today could be responsible for a raging storm next month in New York. In science-speak, this is also known as “sensitive dependence on initial conditions” – and it became the cornerstone of the new field of chaos theory.