Finite Correlation Dimension And Positive Lyapunov Exponents For Surface-Wave Data In The Adriatic Sea Near Venice

We study wind-driven surface wave data taken on an offshore platform in 16 m of water, about 20 km from Venice in the Northern Adriatic Sea. The data are investigated for the effects of chaos and to this end they are subjected to a variety of time series analysis techniques from the field of dynamical systems theory. For certain data sets we find a finite value for the correlation dimension (similar to 7) and a positive value for the largest Lyapunov exponent (similar to 1.5 x 10(-3) bit/sec). In spite of the fact that these results suggest the possibility of chaotic behavior in the data, the correct interpretation is that the data are essentially stochastic, and that the correlation dimensions and Lyapunov exponents result from the anomalous statistical behavior of certain near-Gaussian random processes whose properties we discuss.