Butterfly effects

We have read Tim Palmer’s article “The real butterfly effect and maggoty apples ” (Physics Today, May 2024, page 30) with much interest. He writes that the popular conception of the butterfly effect, in which “the flap of a butterfly’s wings in Brazil can set off a tornado in Texas a week later,” is “folklore” that “isn’t quite correct.”

We recently published a relevant paper on this topic. ¹ We conclude that a butterfly in Brazil cannot cause a tornado in Texas because of its tiny spatial scale and the dominant role of molecular dissipation at that scale.

The notion of a butterfly’s flap causing a tornado is distinct from that of a sensitive dependence on initial conditions affecting the solutions to the equations of motion. Our paper offers a scientific discourse that the former effect is not at all plausible for the real atmosphere. In our chaos studies, ² we have focused on finite predictability in Edward Lorenz’s models of 1963 and 1969 ³ and on three kinds of butterfly effects within those models.

We offer more comments at the online version of Palmer’s article.