Neural networks for blind-source separation of Stromboli explosion quakes

Acernese, F.; Ciaramella, Angelo; Martino, Salvatore; Rosa, R. De; Falanga, M.; Tagliaferri, Roberto

doi:10.1109/tnn.2002.806649

Cited by 41 publications

(34 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It is known as "Strombolian activity". The seismic wavefield of the Stromboli volcano has been investigated, focusing the attention on the explosion quakes (Chouet et al, 1997(Chouet et al, , 1998Del Pezzo et al, 1992;Acernese et al, 2003Acernese et al, , 2004Bottiglieri et al, 2005).…”

Section: Introductionmentioning

confidence: 99%

Statistical analysis of Stromboli VLP tremor in the band [0.1–0.5] Hz: some consequences for vibrating structures

Lauro

Martino

Falanga

et al. 2006

Nonlin. Processes Geophys.

Self Cite

View full text Add to dashboard Cite

Abstract.We analyze time series of Strombolian volcanic tremor, focusing our attention on the frequency band [0.1-0.5] Hz (very long period (VLP) tremor). Although this frequency band is largely affected by noise, we evidence two significant components by using Independent Component Analysis with the frequencies, respectively, of ∼0.2 and ∼0.4 Hz. We show that these components display wavefield features similar to those of the high frequency Strombolian signals (>0.5 Hz). In fact, they are radially polarised and located within the crater area. This characterization is lost when an enhancement of energy appears. In this case, the presence of microseismic noise becomes relevant. Investigating the entire large data set available, we determine how microseismic noise influences the signals. We ascribe the microseismic noise source to Scirocco wind. Moreover, our analysis allows one to evidence that the Strombolian conduit vibrates like the asymmetric cavity associated with musical instruments generating self-sustained tones.

show abstract

Section: Introductionmentioning

confidence: 99%

Statistical analysis of Stromboli VLP tremor in the band [0.1–0.5] Hz: some consequences for vibrating structures

Lauro

Martino

Falanga

et al. 2006

Nonlin. Processes Geophys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…For upgrading our knowledge about the dynamics generating these signals, as a first step, we apply the Independent Component Analysis (ICA) ( [15] and reference therein). It is a well established method, based on Information Theory, to extract, from a recorded signal, independent dynamical systems also non linear, but linearly superposed [16]. The ICA separates also noise.…”

Section: Discussionmentioning

confidence: 99%

Stochastic Resonance Mechanism in Aerosol Index Dynamics

2002

Self Cite

View full text Add to dashboard Cite

We consider Aerosol Index (AI) time-series extracted from TOMS archive for an area covering Italy (7 − 18 o E; 36 − 47 o N ). The missing of convergence in estimating the embedding dimension of the system and the inability of the Independent Component Analysis (ICA) in separating the fluctuations from deterministic component of the signals are evidences of an intrinsic link between the periodic behavior of AI and its fluctuations. We prove that these time series are well described by a stochastic dynamical model. Moreover, the principal peak in the power spectrum of these signals can be explained whereby a stochastic resonance, linking variable external factors, such as Sun-Earth radiation budget and local insolation, and fluctuations on smaller spatial and temporal scale due to internal weather and antrophic components.

show abstract

“…For instance, it was fruitfully applied in volcanic environment [8,13], physics of musical instruments [12,14,15] and dynamical systems in mixtures [16]. Moreover, further studies have been conducted on signals recorded in real environments with delay and reverberation (e.g., convolutive mixtures) [17,18].…”

Section: The Ica Methodsmentioning

confidence: 99%