Wavelet-based filtering of intermittent events from geomagnetic time-series

A four-years-long minute-mean geomagnetic time-series recorded in a middle-latitude observatory was used to investigate the macroscopic dynamics of the magnetosphere–ionosphere complex system. Through probability density and structure function analyses it is shown that the investigated signal exhibits intermittent fluctuations in a certain temporal scale range. This property is regarded as being similar to the turbulent phenomena of magneto-hydrodynamical systems. In this work, discrete orthonormal wavelet transformation and filtering are carried out in order to identify and separate the intermittent parts of the signal from the homogeneous noise-like background. The empirical probability distributions of the laminar time between and the energy content of intermittent events are computed in the time-scale domain representation of the original signal provided by the wavelet transformation. It is shown that the results obtained are in contradiction with each other in the context of the phenomenology of the classical self-organized critical state model. We argue that ‘near-SOC’ or chaotic turbulence model can explain the observed features without ambiguity.