Detection of Nonlinearity and Chaos in Time Series

A method for identification of nonlinearity and chaos in time series is presented. Nonlinearity is tested using a procedure which combines redundancy and surrogate data techniques. After positive identification of the nonlinear character of the data under study, the possible presence of underlying chaotic dynamics can be assessed by a marginal redundancy approach, because of the direct relationship of the marginal redundancy to the Kolmogorov-Sinai entropy of the dynamical system that generates the data.