Measures of complexity in directional change of horizontal wind
by
Amitabha Chanda
Indian Statistical Institute, Calcutta, India
Coauthors: J. Das (ISI, Calcutta, India)

In the present paper we have tried to analyse changes in wind direction in the background of its complexity. Our idea behind choosing wind direction is that whatever be the speed of wind rapid change of direction may be clear manifestation of internal complexity of the dynamical system.

Different measures are available for complex dynamical system. Two most popular measures are coorelation coefficient and mutual information. While correlation coefficient is a well-known statistical method, mutual information is a measure devised on the basis of pattern matching in a stream of data. It has been developed from the information theory. Two other recent methods are fractal/multifractal dimension and epsilon machine reconstruction. In this paper we have described the outline of all these four methods with their pros and cons and general advantages and disadvantages of implementation in analysing complex dynamical systems from time series.

A case study is presented. A data set of horizontal wind direction has been taken for analysis. From the given data we have calculated the change of direction at each step and form a new data set for analysis of complexity. All the four methods have been applied on the data set of direction changes and results obtained. Results show their comparative usefulness as measures of complexity

Date received: November 2, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cafr-90.