He was a meteorologist, but he revolutionized all areas of science, biology to economics through physics. He accidentally discovered the butterfly effect and, as he was brilliant, he realized that he had glimpsed into something very fundamental.
"As a child, I liked doing things with numbers and the weather changes fascinated me," wrote Edward Lorenz in a short autobiography.
Decades later, those two great passions of his - mathematics and meteorology - would come together to change forever the way scientists view natural phenomena.
Edward Lorenz died on April 16, 2008, at the age of 90, of cancer at his home in Cambridge, near Boston. At MIT, where he was professor emeritus and where, in the 1960s, he had discovered the now famous "butterfly effect", which gave rise to no less famous "chaos theory".
The butterfly effect says that there are natural systems whose evolution has nothing to do with chance - for it can be described with perfectly "deterministic" and even relatively simple mathematical formulas - but which are nevertheless totally unpredictable. They are so sensitive to the conditions in which they take place that, starting from almost identical initial conditions, they can evolve into totally different outcomes.
The state of time is one of them, because the atmosphere is a system whose behaviour is extremely sensitive to temperature, humidity, etc., at every instant - in short, to the mere flapping of a butterfly's wings. Scientists then speak of "deterministic chaos," thus bringing together two notions as incompatible as order and disorder.
The idea of a "butterfly effect" was already, at the time, a science fiction theme. In a 1952 tale, A Sound of Thunder, Ray Bradbury imagines how the death of a prehistoric butterfly, when a team of dinosaur hunters travels through time, changes millions of years later the English language, the people and even the results of a presidential election.
And in a novel by Roger Stewart, Storm, with which Lorenz was gifted by his sister when she decided to study meteorology, someone remembers that the sneeze of a man in China could cause a snowstorm in New York. The butterfly effect hung in the air, but nobody had seen it.
The impact of Lorenz's work would not be limited to meteorology; since then, his discovery has touched virtually all areas of the exact and social sciences - and even the study of stock markets.
"By showing that certain deterministic systems have formal limits of predictability, Ed nailed the last nail in the coffin of the Cartesian universe and promoted [chaos theory], which some call the third scientific revolution of the 20th century, after relativity and quantum physics," says Kerry Emanuel, an expert in atmospheric sciences, in a statement issued by MIT.
Despite having received great scientific prizes, such as Crafoord or Kyoto, Lorenz has never had a Nobel Prize.
Born in Connecticut in 1917, Lorenz first studied mathematics at Dartmouth College and Harvard, and in 1948 received a doctorate in meteorology from MIT. And in the early 1960s, he decided to devote himself to trying to make the weather forecast with the help of computers.
It was during the winter of 1961, when Lorenz was doing weather simulations using a simple mathematical model to simulate the Earth's atmosphere, that something happened that, if he hadn't been genius, could have gone completely unnoticed.
One day, as the American James Gleick says in his best-seller Chaos, Lorenz wanted to repeat one of those simulations for longer.
But instead of reusing the initial data from the first simulation, he inadvertently used slightly rounded data, because the computer, which manipulated numbers to six decimal places, printed them with as few as three - and Lorenz, in reintroducing the data to repeat the simulation, used what he had in the computer's "print" instead of the original data. Only the results of the second simulation were totally different from the first.
Lorenz's first reaction was that the computer had some kind of malfunction. But he quickly realized what had happened: it was all because he had changed the "initial conditions" of the simulation.
The reason why the final results of the first and second simulation differed so greatly from a minimum numerical discrepancy (less than 0.1 percent) was not due to a technical error. It was something inherent to meteorological phenomena.
Lorenz published an article in 1963 in the Journal of the Atmospheric Sciences that is now considered a classic. He concluded that "long-term [weather] forecasting is impossible by any method, unless current conditions are exactly known", adding that "since weather observations are inevitably imprecise and incomplete, long-term accurate forecasting does not exist".
That same year, Lorenz gave a lecture at the New York Academy of Sciences on his discovery. And he didn't talk about butterflies but about... seagulls.
He said, "A meteorologist pointed out that if the theory was correct, the flapping of a seagull's wings could change the course of meteorology forever."
His work would, however, be ignored for a decade to come. And it could be said that Lorenz almost owes the consecration of his work... to a butterfly.
It was in December 1972, at the annual conference of the American Association for the Advancement of Science (AAAS) in Washington, DC. "Before the Washington conference," explains Lorenz himself in his 1993 book The Essence of Chaos
"I had sometimes used seagulls as a symbol of sensitivity [to initial conditions]. The passage to the butterfly was actually due to Philip Merilees, who was unable to get in touch with me when he needed to present the conference titles.
Since Lorenz's lecture had no title, Merrilees decided to choose Can the fluttering of a butterfly's wings in Brazil trigger a tornado in Texas? The butterfly did what the seagull had failed to do. The rest is history.