Very long period signals (VLP, 10 s - 30 s)
associated with long period events (0.5 Hz - 5 Hz) were observed at
Etna Volcano, Italy, during June-November 2005. They are only recorded
at the broadband stations nearest to Etna's craters, ECPN, EBEL, EPDN and EPLC.
These stations are part of the permanent seismic
network run by the Catania Section of the
Istituto Nazionale di Geofisica e Vulcanologia (INGV).
Although the signal-to-noise (S/N) ratio for these
VLPs is in general only poor to fair,
they seem to recur, and can be classified into two families.
We improved the S/N by stacking (see 2008 Annual Report), and determined moment tensors for the VLP events using the complete waveform, full moment tensor inversion program (*Minson and Dreger*, 2008).

We calculated both deviatoric and full moment tensors for the VLP stacks of Family I and Family II using the complete waveform inversion code described by *Minson and Dreger* (2008) and synthetic Green's functions for very shallow source depths. The velocity model used to calculate the Green's functions described a simple half-space with a P-wave velocity of 2.0 km/s and a S-wave velocity of 1.2 km/s. A suite of moment tensor inversions was performed at grid points (horizontal spacing 0.25 km; depths in km: 0.25, 0.50, 0.75, 1.0, 1.5) within the volcanic edifice (locations shown in Figure 2.14).
The origin of the rectangular grid was the centroid of the four summit stations. Etna's topography was not included in the calculation of the Green's functions.

For both families, the moment tensor solutions with the best variance reduction (VR)
were in the same region of the edifice as the locations determined for the VLP events using radial semblance (*Cannata, et al.*, 2009). For Family I,
the best solutions had VR 70% and were best explained by sources that are 60-70% isotropic (ISO) (Figure 2.14).
For Family II, they had VR 60% and 60-70% ISO.
Deviatoric solutions for both families had much poorer
VR and waveform fits were clearly less satisfactory.

In moment tensor inversions, the signal to noise ratio (SNR) of the data is clearly important. Fifteen years of experience of moment tensor analysis in California include small events down to M 3.5 and below (*Hellweg et al.*, 2006).
Although events along the central San Andreas Fault are known to be purely double couple (DC), deviatoric moment tensor solutions for small events with low SNR
in the band of analysis (10 s - 50 s) may have up to 30% of their energy modeled by a compensated linear vector dipole (CLVD) mechanism. For the VLP events of Mt. Etna, even the stacks analysed here, the SNR is low.
The DC and CLVD elements of the best solutions vary from grid point to grid point. The eigenvalues describing the deviatoric portions of the solutions also primarily vary randomly in space, with a preference of the largest eigenvector toward a subhorizontal orientation and a slight predominance of SW orientations.
Thus, we are convinced that the deviatoric parts of the moment tensor solution are most likely to be efforts of the inversion to explain the noise.
They cannot be used to interpret the geometry of the source without better data. There is no reason to suppose that a fit including single-force elements (e.g. *Chouet et al.*, 2003, *Chouet et al.*, 2005) would provide greater insight into the source of the VLP events.
On ``source-type'' plots (*Hudson et al.*, 1989), it is notable that all of the moment tensor solutions plot somewhere between ``explosion'' and ``opening crack'' sources.
The scatter gives some sense of the uncertainty in the solutions.

Using Green's functions for full moment tensors calculated using a simple half space velocity model, inversions using the algorithm described in *Minson and Dreger* (2008) indicate that a volume change explains a large portion of the waveforms. We intend to follow up with further analysis to investigate the effects of the simple velocity structure, using Green's functions calculated for the locally used velocity model. We also hope to investigate single source type solutions (i.e. only DC, only CLVD, only ISO), and hope to have longer wavesnippets to improve our understanding of the signal to noise ratio.

Cannata, A., M. Hellweg, G. Di Grazia, S. Ford, S. Alparone, S. Gresta and P. Montalto, Long Period and Very Long Period events at Mt. Etna volcano: characteristics, variability and causality, and implications for their sources, *J. Volcanol. Geotherm. Res*, in press, 2009.

Chouet, B., P. Dawson, T. Ohminato, M. Martini, G. Saccorotti, F. Giudicepietro, G. De Luca, G. Milana and R. Scarpa, Source mechanism of explosions at Stromboli Volcano, Italy, determined from moment-tensor inversions of very-long-period data, *J. Geophys. Res.* *108*, doi:10.1029/20042JB001919, 2003.

Chouet, B., P. Dawson and A. Arciniega-Ceballos, Source mechanism of Vulcanian dagassing at Popocatépetl Volcano, Mexico, determined from waveform inversions of very long period signals, *J. Geophys. Res.* *110*, doi:10.1029/2004JB003524, 2005.

Green, D., and J. Neuberg, Waveform classification of volcanic low-frequency earthquake swarms and its implication at Soufriere Hills Volcano, Monserrat, *J. Volcanol. Geotherm. Res.*, *153*, 51-63, 2006.

Hellweg, M., D. Dolenc, L. Gee, D. Templeton, M. Xue, D. Dreger and B. Romanowicz, Twelve Years and Counting: Regional Moment Tensors in and around Northern California *Seismol. Res. Lett.*, *77(2)*, 221, 2006.

Hudson, J.A., R.G. Pearce and R.M. Rogers, Source type plot for inversion of the moment tensor, *J. Geophys. Res*., *9(B1)*, 765-774, 1989.

Minson, S. and D. Dreger, Stable Inversions for Complete Moment Tensors, in press *Geophys. Journ. Int.*, 2008.

Berkeley Seismological Laboratory

215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760

Questions or comments? Send e-mail: www@seismo.berkeley.edu

© 2007, The Regents of the University of California