Abstract

Abstract. In this paper, spectral and detrended fluctuation analyses, as well as time reversibility and magnitude-sign decomposition, have been applied to the 10-year time-series data resulting from geochemical monitoring of gas emissions on the flanks of Mt. Etna, and gases from a CO2 exploitation well located tens of kilometers from the volcano. The analysis of the time series which showed main effects of fractionation between gases due to selective dissolution in aquifers (e.g., the CO2 concentration series), revealed the occurrence of random fluctuations in time, typical of systems where several processes combine linearly. In contrast, the series of He isotopic composition exhibited power-law behavior of the second-order fluctuation statistics, with values of the scaling exponent close to 0.9. When related to the spectral exponent, this value indicates that the isotopic series closely resemble fractal flicker-noise signals having persistent long-range correlations. The isotopic signals also displayed asymmetry under time reversal and long-range correlation of the associated magnitude series, therefore it was statistically proved the presence of nonlinearity. Both long-range correlation and nonlinearity in time series have been generally considered as distinctive features of dynamic systems where numerous processes interact by feedback mechanisms, in accordance with the paradigm of self-organized criticality (SOC). Thus, it is here proposed that the system that generated the isotope series worked under conditions of SOC. Since the fluctuations of the isotope series have been related to magma degassing, the previous results place constraints on the dynamics of such process, and suggest that nonequilibrium conditions must be dominant. It remains unclear whether the signature of SOC is directly due to volatile degassing from magma, or if it derives from the interaction between melt and the stress field, which certainly influences magma decompression. The strength of scaling appears to increase after 2002 (α values from 0.8 up to 1.2), focusing on transition of the Etnean system from typical SOC toward conditions of lower criticality. By comparing this transition with those of geophysical observables, it can be suggested that the drop in the rate of magma supply, subsequent to the paroxysms of 2001 and 2002–2003, was the main cause of the scaling change.

Highlights

  • Etna system, which suggest it works in conditions of self-organized criticality (SOC); namely, that the system is formed from many interacting components working far from equilibrium, and is characterized by (i) highly nonlinear behavior, (ii) slow driving forces and small perturbations, and (iii) scale invariance of observables in space and time (Sornette and Sornette, 1989; Turcotte, 1992)

  • Irrespective of their location in the volcano edifice, all the sites provide isotope signals marked by power-law exponents close to 1, which are typical of dynamic systems in a self-organized critical state

  • Both TR test and Detrended Fluctuation Analysis (DFA) of the magnitude series, derived from the isotope signals, reveal remarkable degrees of nonlinearity, which are normally found in systems exhibiting complex behavior and feedback mechanisms among interacting processes

Read more

Summary

Introduction

Recent studies of Mt. Etna volcano have analyzed the complex behavior of this system by applying modern statistical tools designed for dynamic systems to geophysical data sets obtained by intensive monitoring of the volcano (Centamore et al, 1997; Vinciguerra and Barbano, 2000; Telesca et al, 2002; Vinciguerra, 2002; Currenti et al, 2005a, b; Walter et al, 2005).

Paonita
Statistical methods
Analysis of unevenly sampled time series
Detrended fluctuation analysis
DFA in short time series
Magnitude and sign decomposition
Reliability of the statistics: the method of surrogate series
DFA and magnitude-sign decomposition on interpolated series
10 White RSS
Reliability of TR tests
Application to geochemical time series
Time series of He and CO2 concentration
Summary and conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call