Application of the Wiener filter to magnetic profiling in the volcanic environment of Mt. Etna (Italy)

Abstract

The frequency-domain Wiener filtering was applied to magnetic anomalies in the volcanic area of Mt. Etna. This filter, under suitable conditions (additive noise, linear processing and mean-square error criterion), can furnish an effective tool for discriminating the geologic feature of interest (the signal) from the noise. The filter was first tested with synthetic data. Afterwards it was applied to a magnetic profile carried out across the principal fault system of the Mt. Etna volcano, that hosted the dykes feeding both the 1989 and the 1991-93 eruptions. The magnetic anomalies linked to the volcanic section and those linked to the contact between the clay basement and the lava coverage show significant spectral overlap. Thus by estimating the power spectrum of the signal, obtained resolving the forward problem, a least-squares Wiener filter has been designed. In such context, it was possible to verify the effectiveness of Wiener filters, whereas traditional band-pass filtering proved inadequate. In fact, analysis of the noise showed that all the meaningful components of the observed magnetic field were resolved. The results put further constraints on location and geometry of the shallow plumbing system of Mt. Etna.