
A set of states in a dynamical system towards which it evolves, exhibiting sensitive dependence on initial conditions.A chaotic attractor is a geometric shape that describes the long-term behavior of a dynamical system that is both deterministic and unpredictable in detail. In such a system, the state evolves over time, and no matter where it starts, it eventually settles onto this attractor—a set of states it keeps visiting. However, the path it takes on the attractor is exquisitely sensitive to tiny differences in starting conditions, so two nearly identical initial states quickly diverge and follow completely different trajectories.
The concept emerged from Edward Lorenz's work in the 1960s on simplified atmospheric convection, where he discovered that a three-variable model produced a butterfly-shaped attractor with chaotic behavior. The term 'strange attractor' was coined by David Ruelle and Floris Takens in 1971.
The attractor is formed by the system's equations, which pull the state toward certain regions of its state space while also stretching and folding the trajectories like a baker kneading dough. This stretching amplifies small uncertainties, while the folding keeps the motion bounded, so the system never escapes but also never repeats exactly. The result is a fractal structure that is neither a simple point, a closed loop, nor a smooth surface, but a strange, infinitely detailed object.
Chaotic attractors are used to analyze turbulence, design secure communication systems that exploit chaos, and understand the limits of predictability in weather and climate.
The Lorenz attractor, the first chaotic attractor studied, is shaped like a butterfly's wings, giving rise to the famous 'butterfly effect' metaphor.











