Lorentz was modeling weather with a system of three coupled non-linear differential equations. He found that very small changes in system initial conditions caused wildly divergent behavior of the system state as the system equations were solved further out in time. He proposed his findings as a reason for the poor results in making predictions of the weather.

Thompson was skeptical and used a “discretized noisy version” of the original model to test Lorenz’s conclusions. He found that the addition of a small amount of noise resulted in stable and repeatable results rather than chaotic behavior. In this case, chaos was a probably a mathematical artifact and the addition of noise had a stabilizing effect on system behavior.

Thompson’s conclusion? Physical processes such as weather are driven by large effects such as El Nino and jetstream currents. When realistically modeled they don’t exhibit chaotic behavior, so the Butterfly effect probably isn’t real.