Abstract

Abstract. We analyze 25 episodes of volcanic tremor recorded from 22 November until 31 December 2009 at Mt Semeru volcano in order to investigate their spectral and dynamical properties. The overtone frequencies for most of the tremor events indicate a pattern of period-doubling, which is one possible route that can lead a system to chaotic behavior. Exponential divergence of the phase space orbits is a strong indicator of chaos and was quantified by estimating the maximal Lyapunov exponent (MLE) for all tremor events. MLEs were found to vary linearly with the number of frequency overtones present in the tremor signals. This implies that the tremor source at Semeru fluctuates between a quasi-periodic state with few overtone frequencies (2–3) and small MLEs (~0.013), and a chaotic one with more overtones (up to 8) and larger MLEs (up to 0.039). These results agree well with the tremor generation model suggested previously by Julian (1994), which describes wall oscillations of a crack excited by unsteady fluid flow. In this model, as fluid pressure increases, a period-doubling cascade leads to numerous new frequencies and a chaotic tremor signal. The temporal variation of MLEs exhibited significant fluctuations from 23 until 31 December when the eruptive activity shifted from explosive to effusive. Such a situation may reflect variable fluid pressure conditions inside the conduit, where at first magma is accumulated and subsequently is erupted, releasing the buildup of pressure. Our results give further evidence for the role of nonlinear deterministic processes in generating volcanic tremor and call for similar investigations to be conducted in other volcanoes.

Highlights

  • A key issue that is highlighted during the analysis of time series is what kind of process is generating this observed signal

  • The main conclusions of this work can be summarized as follows: 1. Volcanic tremor signals recorded at Mt Semeru during November–December 2009 consist of 2–8 overtone frequencies and a fundamental frequency between 0.5 and 2 Hz without exhibiting gliding effects

  • The overtone frequencies for some of the tremor events indicate a pattern of period-doubling, which is one possible route to chaotic behavior

Read more

Summary

Introduction

A key issue that is highlighted during the analysis of time series is what kind of process is generating this observed signal. Several studies have applied nonlinear time series analysis methods to tremor recorded at volcanoes worldwide in order to detect the presence of deterministic structure (Chouet and Shaw, 1991; Godano et al, 1996; Carniel, 1996; Carniel and Di Cecca, 1999; Konstantinou, 2002; Konstantinou and Lin, 2004; De Lauro et al, 2008; Cannata et al, 2010) Their results revealed the existence of a strange attractor that describes the tremor source process in phase space and is characterized by a relatively small number of degrees of freedom. The maximal Lyapunov exponent (thereafter called MLE) is a strong indicator of chaos and is estimated with the aim of investigating its possible variations and relationship with the physical mechanism that is causing tremor

Data and spectral analysis
Estimation of embedding parameters
Estimation and characteristics of MLEs
Findings
Discussion
Conclusions
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