On the butterfly effect

Alexander Shnirelman

DOI: https://doi.org/10.48550/arXiv.1605.08271

2016-05-23

Abstract:The term "butterfly effect" means an extreme sensitivity of a dynamical system to small perturbations: "The beating of a butterfly wing in South America can result in the considerable change of positions and force of a tropical cyclon in Atlantic 2 weeks later". Numerical simulations of R.Robert show the absence of the butterfly effect in some simple flows of 2-d ideal incompressible fluid which is a model of the atmosphere. In this work a more complicated flow is considered. Numerical simulation demonstrates the butterfly effect in the strongest form. The effect is robust, and the experiment is 100% reproducible.

General Physics

What problem does this paper attempt to address?

This paper attempts to explore and verify the existence of the "butterfly effect." Specifically, the author studies how small perturbations in initial conditions can lead to significant changes in a hydrodynamic system by numerically simulating the Euler equations on a 2-dimensional torus. The main objectives of the paper are:

1. **Verify the "butterfly effect"**: Through numerical simulations, verify whether small changes in initial conditions (such as the "flapping of a butterfly's wings") can lead to significantly different outcomes in the system after a period of time.
2. **Study fluid instability**: Investigate strong instabilities in fluid dynamics, particularly different types of flow behaviors and their sensitivity to initial conditions.

The author verifies the "butterfly effect" by setting two very close initial conditions and observing the differences in their evolution over time. The results show that even with very small differences in initial conditions, the final flow field configurations can be significantly different, thus proving the existence of the "butterfly effect."