Functioning of Piton de la Fournaise volcano inferred from continuous seismological observations

Functioning of Piton de la Fournaise volcano inferred from continuous seismological observations

Piton de la Fournaise volcano, La Réunion

F. Brenguier (1), N. Shapiro (1), M. Campillo (2), V. Ferrazzini (3), Z. Duputel (3), O. Coutant (2), E. Rivemale (1), and A. Nercessian (1)

2

1

Laboratoire de Géophysique Interne et Tectonophysique

Institut de Physique du Globe de Paris

3Observatoire Volcanologique du Piton de la Fournaise, IPGP, La Réunion, France

Piton de la Fournaise

Sea leve

l

S 21°

E 55°30

Scheme

Topographic map of La Réunion island

Seismic activity and volcanic eruptions

Volcanic eruptions

Cum

ula

tive

sei

smic

ity

days

Explore transient periods from continuous seismological observations.

Waveform recognition

Waveform recognition (work from Elodie Rivemale, Master student)

Calculate correlation coefficient between the master event and the raw seismic signal

Master event : volcano-tectonic event located near the magma reservoir (sea level)

Cor

r. C

oeff

.

Waveform recognition - Results

Waveform recognition

* We identify a constant seismicity rate within transient periods that could berelated with the pressurization of the magma reservoir.

* Comparison of eruptions ER4 and ER5: reactivation of similar dykes

Cum

ula

tive

sei

smic

ityC

umu

lativ

e s

eism

icity

days

Classical Automatic detection

Passive seismic noise monitoring

Seismic noise energy is quite uniform at frequencies [0.1 - 1] Hz – Oceanic originimplies good azimuthal coverage.

Using Seismic noise ?

Green’s function reconstruction from seismic noise

noise sources

receivers

Yang et al. (2007) Shapiro et al. (2005)

Europe South California

PBRZ NCR

Internal structure

3D surface wave tomography using correlations of seismic noise

Brenguier et al., GRL, 2007

Internal structure

solidified dykeseffusive material

3D surface wave tomography using seismic noise

Passive seismic imaging

How the velocity structure evolves along time ?

Time 1 Time 2

time evolution

Passive seismic imaging

Measuring a uniform relative velocity perturbation(Known as Moving Cross-Spectrum Window Method or Coda-wave interferometry)

Synthetic velocity decrease

Data processing

Measuring relative velocity perturbations from observed noise cross-correlations

Results

Testing the method with data from 1999-2000

Brenguier et al., Nature Geoscience, 2008

Time dependent regionalization

Time dependent regionalization

Magmatic intrusive complex

Toward real-time monitoring

Eruption of July 2006

http://www.fournaise.info/eruption2avril07.php

The last eruption of April 2007

Toward real-time monitoring

Toward real-time monitoring

Link between velocity change maxima and emmited volumes

Conclusions

We measure relative velocity changes with a precision of 0.02 %,

These changes are linked to dilatation induced by stress changes,

We identified precursors to the volcanic eruptions.

Prospects

We are looking to localize the velocity changes at depth (4D tomography),

Increase the time resolution,

Conclusion

Aknowledgments

Piton de la Fournaise observatory staff and C. Sens-Schoenfelder, L. Stehly, P. Gouédard, P. Roux, G. Poupinet.

Thank you ! – Collaboration with ERI: monitoring Mt Asama volcano

Regionalization procedure

Relative time perturbations for one receiver pair

5 days predated sliding window

5 days predated sliding window

Peltier et al. (2005)

Peltier et al. (2006)

Monitoring ground deformations

Long term variations (few months) Short term variations (few hours)