Adaptive Local Iterative Filtering: A Promising Technique for the Analysis of Nonstationary Signals

Many real-life signals and, in particular, in the space physics domain, exhibit variations across
different temporal scales. Hence, their statistical momenta may depend on the time scale at which the signal
is studied. To identify and quantify such variations, a time-frequency analysis has to be performed on these
signals. The dependence of the statistical properties of a signal fluctuation on the space and time scales
is the distinctive character of systems with nonlinear couplings among different modes. Hence, assessing
how the statistics of signal fluctuations vary with scale will be of help in understanding the corresponding
multiscale statistics of such dynamics. This paper presents a new multiscale data analysis technique, the
adaptive local iterative filtering (ALIF), which allows to describe the multiscale nature of the geophysical
signal studied better than via Fourier transform, and improves scale resolution with respect to discrete
wavelet transform. The example of geophysical signal, to which ALIF has been applied, is ionospheric radio
power scintillation on L band. ALIF appears to be a promising technique to study the small-scale structures
of radio scintillation due to ionospheric turbulence.