If you weren't able to make it to Monday's MIT Atmospheric Science Seminar, Prof. Palmer has generously made his presentation The "Real" Butterfly Effect available.
The phrase “The Butterfly Effect” is almost universally used to describe sensitive dependence on initial conditions in chaotic systems (be they high or low order). However, this is not what Lorenz originally had in mind by this phrase. Rather he postulated the existence of something much more radical: that a high dimensional system like the atmosphere may have a finite predictability horizon which cannot be extended in time, no matter how small the initial uncertainties are. Is there evidence for “The Real Butterfly Effect” in the real world, and is “The Real Butterfly Effect” a property of the Navier-Stokes equations? In this seminar, I will review some of these issues and then conclude that an understanding of the “The Real Butterfly Effect” is of crucial practical importance as we aim to provide reliable weather and climate predictions to a range of real-world applications from health to agronomy to hydrology.
Thanks to Tim Palmer for sharing his presentation.