Low-dimensional chaos in a hydrodynamic system

Evidence is presented for low-dimensional strange attractors in Couette-Taylor flow data. Computations of the largest Lyapunov exponent and metric entropy show that the system displays sensitive dependence on initial conditions. Although the phase space is very high dimensional, analysis of experimental data shows that motion is restricted to an attractor of dimension 5 for Reynolds numbers up to 30 percent above the onset of chaos. The Lyapunov exponent, entropy, and dimension all generally increase with Reynolds number.