A wavelet-based mode decomposition
University of Liège, Department of Mathematics, 12, Grande Traverse Bât. B37, 4000 Liège, Belgium
Corresponding author: a S.Nicolay@ulg.ac.be
Revised: 28 January 2011
Published online: 9 March 2011
We propose a wavelet-based method for analyzing non-stationary data. The idea, inspired by the empirical mode decomposition, is to decompose a data set into a finite number of components, well separated in the time-frequency plane, plus a residue, such that each component has a zero mean and is associated to one frequency only. When applied to climatic data, this method gives interesting results.
© EDP Sciences, Società Italiana di Fisica, Springer-Verlag, 2011