Edward N. Lorenz ScD ’48, an MIT meteorologist whose meticulous attempt to predict the weather through an early computer unraveled into what became known as the chaos theory, died of cancer at his home in Cambridge Wednesday. He was 90.
His development of chaos theory is considered a seminal moment in 20th century science. Also called “deterministic chaos,” the theory brought about “one of the most dramatic changes in mankind’s view of nature since Sir Isaac Newton,” said the committee that awarded Dr. Lorenz the 1991 Kyoto Prize for basic sciences.
The practical effect of this theory was delivered most starkly in an academic paper he offered in 1972: “Predictability: Does the flap of a Butterfly’s Wings in Brazil set off a Tornado in Texas?”
The phrase “butterfly effect” would become part of the lexicon of both pop science and pop culture.
Dr. Lorenz suggested the answer to the question he raised was a possible “yes,” and for decades he contended such variables limited the accuracy of long-term weather forecasts, even with the most sophisticated computerized systems.
In a statement, MIT said Dr. Lorenz’s insight led to the conclusion that it might be fundamentally impossible to predict weather beyond two or three weeks with a reasonable degree of accuracy.
It was an assessment in which Dr. Lorenz reluctantly agreed. “There’s little hope for ever predicting weather a month ahead,” he told The Boston Globe in 1984.
His work on the theory began by accident in 1961. He was trying to determine how accurate a computer could predict long-term weather patterns. He ran one simulation with a computer model, then wanted to extend the forecast, so he added a second simulation, with the same parameters and conditions of the first model. The weather pattern should have seamlessly flowed into the second simulation.
Instead, the trajectories radically shifted.
The problem: a rounded decimal number. Dr. Lorenz realized that the computer stored numbers to an accuracy of six decimal places but, to save space, printed out results shortened to three decimal places. So, for example, 0.310625 became 0.311. For the second simulation, he had used the shortened figure.
Even this minute discrepancy drastically altered the forecast.
Tiny changes, in effect, could have catastrophic, and often unpredictable, consequences. And they made perfect predictions of weather, even through the emerging power of computers, impossible: Exact measurements of all the conditions could be upset by one small event, such as the flap of a gossamer wing.
The development of this theory changed not only how scientists viewed the prediction of weather, but also had applications in such sciences as fluid dynamics.
In his desire to improve weather forecasts, Dr. Lorenz had developed the theory that, ultimately, undermined the possibility of flawless predictions.
“Any imprecision will throw you off,” Dr. Lorenz told the Globe. “Even the thickness of the pen line on a graph can be enough.”
Dr. Lorenz’s interest in the weather was influenced by a life-long love of nature. He wrote in an autobiographical sketch, “As a boy I was always interested in doing things with numbers, and was also fascinated by changes in the weather.”
Born in West Hartford, Conn., in 1917, Mr. Lorenz often vacationed with his family in New Hampshire’s White Mountains, where they had a summer home. He would return to these mountains the rest of his life.
“He was like a mountain goat,” said Kerry Emanuel ’76, a professor of atmospheric science at MIT and friend of Dr. Lorenz, who sometimes went hiking with him. “He knew every trail in the White Mountains and the Rockies.”
Despite his quiet nature, he had a surprising sense of humor while on the trail.
“I remember one small hike in a desert in southern California, there was a big pit of sleeping coyotes,” Emanuel said. “All of a sudden I heard a yelp and I jumped and looked around. It was Ed. He learned to imitate a coyote.
“And then all the other coyotes began yelping. He was full of surprises.”
Dr. Lorenz held degrees from Dartmouth College, Harvard University, and MIT. He also served as a weather forecaster for the US Army during World War II, deciding then to do his graduate work at MIT.
Mr. Lorenz worked at MIT from 1948 to 1987 when he retired as an emeritus professor. In 1983, he and his colleague Henry Stommel shared a $54,000 Crafoord Prize from the Swedish Academy of Sciences, an international prize considered to be nearly equal to a Nobel Prize. They received it for their research that broadened the understanding of the atmosphere and the sea.
For all his honors, however, Dr. Lorenz was a humble scientist. Emanuel said.
“He was a very unassuming man,” he said.
Dr. Lorenz’s wife, Jane, died in 2001. He leaves three children, Nancy, of Roslindale, Edward H. Lorenz ’75 of Grasse, France, and Cheryl, of Eugene, Ore.; and four grandchildren.
Material from The New York Times News Service was used in this report.