Bad Forecast? Blame the Butterflies
Farewell to Lorenz, the man behind the "butterfly effect"
By Steve Tracton
"Don't blame us. Blame the butterflies" -- maybe that's what meteorologists should say when they miss a forecast.
One person who most certainly would have appreciated such an excuse is Ed Lorenz. It was Lorenz who was one of the first to describe what is known as the "butterfly effect," which conveys the more esoteric concept of "sensitive dependence on initial conditions."
In a nutshell, the idea is that the flapping of a butterfly's wings over Mexico, for example, might cause a sequence of events leading to a snowstorm (or some other type of weather) at a later time in, let's say, Washington, D.C.
As anyone who knew or knew of Ed Lorenz, I was deeply saddened by his passing last week.
Ed is probably best known for his pioneering work on "chaos theory" and its relevance to issues of atmospheric predictability. He was one of the few real giants in contributions to knowledge and understanding of the atmosphere. As true for most of his students and colleagues at MIT, he was to me not just a teacher and mentor, he was a friend.
During my graduate school days in the late 1960s, my office was adjacent to Ed's. Aside from teaching class and counseling thesis candidates, he spent much of his time number-crunching the impact of butterflies flapping their wings (i.e., miniscule differences in initial conditions) on his simplistic, but realistic computer models of the atmosphere.
However, I was largely ignorant of the nature, let alone importance, of this work until almost too late. Too late in the sense of not being prepared for a related question he might -- and ultimately did -- have on the "General Exam," which one had to pass before proceeding on a Ph.D. track. Fortunately, a fellow student gave me the "heads up" in the nick of time.
Or maybe unfortunately -- like many who have gone down the difficult path of a Ph.D., there were moments I wish I had flunked the exam and wanted to punch the person who alerted me about Ed's work.
Either way, perhaps that alert was the butterfly that forever changed the course of my career and life in general.
As for real-life butterflies, millions of monarchs are now migrating from their winter haunts in Mexico toward the northern United States and Canada. You can actually keep track of the migration here.
Just to ensure they wreak as much havoc as possible on predictability, the monarchs, whose life span is far too short to travel thousands of miles, sequentially spawn two to four generations of new butterflies as the migration moves northward.
The migration begins in the February-March period. The second generation is born -- via the life cycle of egg, caterpillar, chrysalis and adult butterfly -- in March and April, with the third appearing in the July-August timeframe. And not much later, in the fall, a fourth generation begins its way south for the winter.
So, for the better part of the year, perhaps real butterflies are to blame for forecast busts. That might be more believable than the usual refrain - the models blew it!
The author is the current chair of the D.C. chapter of the American Meteorological Society.
The comments to this entry are closed.