HIGH PRECISION LOCATIONS OF LP EVENTS ON MT. ETNA: RECONSTRUCTION OF THE FLUID-FILLED VOLUME S. GAMBINO Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Catania, P.zza Roma 2, 95123 Catania, Italy (gambino@ct.ingv.it) Received: December 18, 2002; Revised: January 27, 2006; Accepted: March 4, 2006 ABSTRACT During 1991−93 at Mount Etna, long-period (LP) events occurring in swarms characterized the evolution of the eruption. The presence of multiplets i.e. groups of events with similar waveform signatures, has been recognized within this activity. Traditional techniques for locating LP events do not allow obtaining reliable hypocenters, which have only succeeded in placing earthquakes in a roughly 1 km2 area slightly east of the Mt. Etna Northeast Crater. Hypocenters have been relocated in two steps: the absolute location has been improved using Thurber’s code and a complex 3D velocity model; a highly precise relative location has been applied on multiplets to define the source geometry. 3D locations and high precision analysis suggest that during the 1991−93 eruption the resonator producing LP events was a part of the uppermost Northeast Crater conduit, measuring 210 meters in height and 45−50 meters in diameter. K e y w o r d s : waveform correlation, stacked events, 1991−93 eruption, conduct geometry, Mt. Etna Northeast Crater 1. INTRODUCTION Mount Etna is a basaltic volcano situated on the eastern coast of Sicily (Fig. 1) and is one of the most active volcanoes in the world. The volcanic edifice has a basal diameter of about 40 km, is just over 3200 m high and has four active summit craters (Fig. 1a). Located between the compressive domain of Western-Central Sicily and the tensional domain of the Calabrian Arc, the volcano has formed at the intersection of two regional fault systems, having NNW-SSE and NE-SW trends (e.g. Lo Giudice et al., 1982; Monaco et al., 1997), respectively (Fig. 1b). Activity on Mt. Etna may be divided into two types: lateral flank eruptions occurring along fracture systems and persistent activity that comprises phases of degassing alternating with Strombolian activity, which may occasionally evolve into lava fountains and effusive activity. Stud. Geophys. Geod., 50 (2006), 663−674 © 2006 StudiaGeo s.r.o., Prague 663 S. Gambino Fig. 1. Map of the 1992 permanent seismic network. Solid circles and triangles indicate 1 component (vertical) and 3 component stations respectively. The two insets show (a) the summit area and (b) the main regional fault systems. Seismicity at Mt. Etna is usually defined as seismo-tectonic activity and the observed seismic signals can generally be divided in two major groups (Patanè et al., 2004): volcanic-tectonic (VT) earthquakes generated by tectonic stress and/or by stress arising from rising magma; and shocks linked to fluid dynamics, i.e. long-period (LP) events (Chouet, 1985), and volcanic tremor. Analysis of LP events is of particular interest to understand eruptive processes as their source mechanism is believed to directly involve fluid transport (Chouet, 1988; Battaglia et al., 2003). However, because waveforms are characterized by emergent first arrivals, LPs are difficult to locate with conventional techniques. It is not unusual to find groups of seismic events with very similar waveforms. They are commonly called multiplets and generally represent the result of stress releases along the same structure generated by very closely located sources (Pechmann and Kanamori, 1982). The application of cross-spectral techniques (Poupinet et al., 1984; Frèmont and Malone, 1987) and cross-correlation methods (Haase et al., 1995) permits achieving a precision less than the digitations interval finding relative arrival time shifts of about 664 Stud. Geophys. Geod., 50 (2006) High Precision Locations of LP Events on Mt. Etna: Reconstruction of the Fluid-Filled Volume 1−3 ms at the different stations. This relative time precision allows relocating the earthquakes with accuracy within 5−20 m. In tectonic areas, these kinds of analyses allow the reconstruction of seismogenic structures and the discrimination of the fault plane from the auxiliary in the focal mechanism solution (e.g. Cattaneo et al., 1997). The application of these methods on different volcanoes (e.g. Got et al., 1994; Jones et al., 2001; Baher et al., 2003) and recently also on Mt. Etna (Alparone and Gambino, 2003; Brancato and Gresta, 2003) has generally been used to relocate VT events, highlighting their possible relationship with volcanic activity. Precise relocations of LP events are rarer and have been performed on Kilauea Volcano (Wolfe et al., 2003; Battaglia et al., 2003), Soufreire Hills Volcano Montserrat (Rowe et al., 2004) and Redoubt Volcano Alaska (Rowe, unpublished data). In this work, a cross-spectral technique (Fremont and Malone, 1987) has been applied to a subset of 16 long-period events, occurring in April 1992 and located near the Northeast summital Crater (NEC) of Mt. Etna volcano, which showed nearly identical seismic waves. LP seismicity is generally considered as the resonance of a fluid filled volume in response to a trigger (e.g. Chouet, 1985, 1996). In this paper, accurate relative locations for two LP families recorded in 1992 has enabled obtaining a picture of the shape of the resonance volume. 2. GENERAL FEATURES OF THE 1991−93 ERUPTION The 1991-93 eruption began on December 14, 1991 and continued until March 31, 1993. It represents one of the most prolonged and volumetrically significant events occurring at Mt. Etna over the last 300 years (Fig. 1). The lateral eruption started from a fracture system originating at the base of the SouthEast Crater (SEC) and propagated to the SSE down to 2200 meters elevation along the western wall of the Valle del Bove (Barberi et al., 1993). The eruption was characterized by a low-explosive activity but produced the largest lava emission forming a lava field of about 7.6 km2 with a total volume estimated of 235 × 106 m3 of magma (Barberi et al., 1993; Calvari et al., 1994). The eruptive activity was characterized by an effusion of magma accompanied by Strombolian explosions at the end of the fracture system. This Strombolian activity ceased in March 1992 and thereafter eruptive activity was limited to a sustained lava-flow from the distal end of the fracture system. In spite of its importance, the eruption was preceded by a relatively short (few hours) and small (about 250 earthquakes) seismic sequence. These earthquakes were located at high elevations beneath the southern-flank of the volcano, near to where eruptive activity started (Ferrucci and Patanè, 1993; Chouet et al., 1994). The fracturing activity lasted some days, and during the eruption only a few minor swarms occurred. On the contrary, volcanic tremor was recorded during the entire episode of eruptive activity. Beginning in January, 1992 numerous LP events were recorded until the end of the eruption (Falsaperla et al., 2002). These events usually occurred in swarms in which Chouet et al. (1994) recognized the presence of multiplets. Stud. Geophys. Geod., 50 (2006) 665 S. Gambino 3. DATA Mt. Etna seismic activity is recorded by the Istituto Nazionale di Geofisica e Vulcanologia (INGV) which currently operates a permanent local seismic network composed of 50 seismic stations (Patanè et al., 2004). All data are transmitted in real time by means of radio links and telephone cables to the data acquisition center in Catania. The seismicity here analyzed was recorded in 1992 when the permanent seismic network comprised nine stations (Fig. 1). The network included seven analog stations equipped with vertical 1-Hz seismometers, and two digital stations (ESP and EGA) featuring broadband (0.1−100 Hz) three-component sensors. The seismic signals, transmitted either by cable or radio from the remote sites to Catania, were recorded with sampling at 125 Hz and anti-aliasing filter at 35 Hz. a) b) c) Fig. 2. (a) Seismograms, (b) power spectra and (c) spectrograms of the vertical ground velocity for the #2 LP event (Table 1) recorded at PDN, ESP and CTS stations. 666 Stud. Geophys. Geod., 50 (2006) High Precision Locations of LP Events on Mt. Etna: Reconstruction of the Fluid-Filled Volume Beginning in January, 1992, thousands of low-magnitude LP events characterized the seismicity of the 1991-93 eruption. These LP events exhibit emergent P phases, weak or non-existent S phases, and a spectral band ranging between 0.5−9 Hz with dominant frequencies between 2−4 Hz (Fig. 2). 1992 LP events are characteristic of those occurring throughout the entire 1991−93 Etna eruption. They generally occurred in swarms that were temporally correlated with several visual observations of repeated collapses of the crater floor of the NEC which led to the formation of a pit crater (Calvari et al., 1994; Coltelli et al., 1998). Within this seismicity, Chouet et al. (1994) and Falsaperla et al. (2002) distinguished four families of LP events (F1, F2, F3, F4) based on similarities between waveforms (Table 1). The relocation analysis requires that events are recorded at least 5 stations; for this reason the 16 events belonging to F1 and F3 have been used for analysis, whereas the events of F2 and F4 have been rejected owing to recording problems at some stations. It is noteworthy that events of F2 are very similar to F1 and F3, while F4 is composed of only 3 events (Falsaperla et al., 2002; Table 1). The 16 earthquakes have been located by using the HYPOELLIPSE program (Lahr, 1989) which accommodates the difference in altitude of the seismic stations with a onedimensional velocity crustal model derived from Hirn et al. (1991) and the reference plane at 2900 m a.s.l. The uncertainties in the P-phase picks causes a scattered distribution of the LP hypocenters, so to improve the signal-to-noise ratio and reduce location errors the stacked traces (Fig. 3) of the two families have been considered. These fictitious events have been located and used to aid the picking of the emergent P-phases of the distant stations in similarity to Rowe et al. (2004). The epicenters are enclosed in a roughly 1 km2 area slightly east of NEC and the focal depths scatter in a range between 0.17 and 0.76 km above sea level (a.s.l.); location parameters reported in table 2 show horizontal and vertical errors ≤ 0.8 km, azimuthal gap comprised between 89° and 135° and RMS < 0.10 s. However, HYPOELLIPSE program is based on the Geiger method and the error estimates are only formal values and do not provide reliable information on the true uncertainty in the hypocenter; moreover, Patanè et al. (2003) showed that the pronounced lateral heterogeneity of the crust beneath Mt. Etna strongly influences earthquake locations. This motivated the relocation of the seismicity using the SIMULPS14 code (from Thurber, 1993) and the 3D velocity model of Aloisi et al. (2002). 3D analysis has improved locations; in fact epicenters (Table 3 and Fig. 4) are closer to NEC crater, shallower (depth between 1.66 and 2.29 km a.s.l.) and show a lower RMS (SIMULPS14 do not furnish horizontal and vertical errors). Table 1. Events and time interval of the four LP families. Family N of events 1 2 3 4 8 11 8 3 Stud. Geophys. Geod., 50 (2006) Time of first event 02/04/92 12:48 06/09/92 21:26 15/04/92 02:04 25/07/92 21:25 Time of last event 03/04/92 13:58 30/09/92 05:13 15/04/92 17:28 07/09/92 03:37 667 S. Gambino Fig. 3. Correlated waveforms, recorded at PDN station, belonging to the families 1 and 3. 4. RELOCATION METHOD Relocation of multiplets was performed using a cross-spectrum method based on that discussed by Fremont and Malone (1987). This method permits very accurate relative timing (dt) for pairs of earthquakes with very similar waveforms (doublets) and successively to perform precise relative relocations. Each doublet comprises a reference event (master event) and one of the other events belonging to the same multiplet (Moriya et al., 2003; Battaglia et al., 2004). The differences dt in P arrival time between seismograms from the same station of a doublet have been computed in the frequency domain using the phase of the cross spectrum obtained on a short window containing the whole P phase. In particular, dt is proportional to the slope of the phase of the cross spectrum which can be written Φ(f) = 2πdtf, plotted versus the frequency (f). The degree of success of this procedure depends on the similarity of the waveforms and the signal to noise ratio on all traces. A parameter measuring the similarity degree between two waveforms is the coherency C(f), defined as the ratio of cross-spectrum modulus over the product of the spectra of the two signals (Fremont and Malone, 1987): C ( f ) = γ s1s 2 ( f ) ⎡ γ s1 ( f ) γ s 2 ( f ) ⎤ , ⎣ ⎦ 668 Stud. Geophys. Geod., 50 (2006) High Precision Locations of LP Events on Mt. Etna: Reconstruction of the Fluid-Filled Volume where γ s1 ( f ) = S1 ( f ) S1∗ ( f ) and γ s 2 ( f ) = S2 ( f ) S2∗ ( f ) are the spectra of the first and second signal, respectively, and γ s1s 2 ( f ) = S1 ( f ) S2∗ ( f ) is the cross-spectrum (* denoting complex conjugate). Mean coherency calculated above the P-wave frequency interval defines the quality (Qw) of comparison between two waveforms. Two very similar signals have a Qw > 90. A decreasing value of Qw indicates a poorer similarity and generally 80 is the threshold below which it is difficult to obtain dt with acceptable errors (Fremont and Malone, 1987). The position and the difference in origin time of the second with respect to the master event have been determined using a decomposition in singular values technique discussed by Aki and Richards (1980). The parameters needed to obtain a suitable relocation of an event are: the P-wave velocity in the source volume, the take-off angles and azimuths of stations from the reference event and a minimum number of 5 time differences. 5. TIME DIFFERENCES AND RELATIVE RELOCATIONS The set considered comprises 16 long-period events (Fig. 3) belonging to F1 and F3. For each event we considered the waveforms of the 7 functioning stations that permit an azimuth coverage with a gap of ca. 90° (Table 2). Table 2. Hypoellipse location parameters of the analyzed earthquakes. n Date 1 Stacked 2 920402 3 920402 4 920403 5 920403 6 920403 7 920403 8 920403 9 920403 10 Stacked 11 920415 12 920415 13 920415 14 920415 15 920416 16 920416 17 920416 18 920416 Origin time Latitude Long. event of F1 12:48:12.56 21:38:26.83 03:06:05.04 07:32:07.23 08:33:32.51 10:18:15.41 12:19:32.29 13:58:01.46 event of F3 02:04:40.02 02:16:17.37 02:45:15.24 03:06:13.92 02:51:16.11 05:26:49.27 17:20:02.44 17:28:35.34 37.7547 37.7550 37.7532 37.7527 37.7531 37.7532 37.7528 37.7533 37.7547 37.7547 37.7550 37.7543 37.7530 37.7545 37.7542 37.7547 37.7548 37.7546 15.0125 15.0083 15.0105 15.0132 15.0120 15.0101 15.0110 15.0088 15.0075 15.0071 15.0068 15.0073 15.0082 15.0080 15.0080 15.0072 15.0075 15.0075 Depth −0.31 −0.17 −0.36 −0.30 −0.33 −0.36 −0.34 −0.29 −0.54 −0.55 −0.53 −0.51 −0.76 −0.54 −0.51 −0.51 −0.50 −0.51 Gap No RMS ERZ ERH Q 133 131 134 135 133 134 132 131 133 93 94 92 89 91 92 92 93 92 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 0.05 0.06 0.07 0.07 0.09 0.07 0.07 0.08 0.05 0.03 0.04 0.03 0.06 0.05 0.06 0.05 0.07 0.06 0.6 0.6 0.6 0.8 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.6 0.6 0.6 0.7 0.6 0.5 0.6 0.5 0.6 0.7 0.5 0.5 0.6 0.5 0.5 0.6 0.5 0.5 0.5 0.6 0.5 0.6 0.5 B B B B B B B B B B B B B B B B B B GAP - azimuthal gap (degrees); RMS - travel-time residual root mean square (s); No - number of P arrivals. Stud. Geophys. Geod., 50 (2006) 669 S. Gambino Table 3. 3D location parameters of the analyzed earthquakes. N Date Origin time Latitude Longitude 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Stacked 920402 920402 920403 920403 920403 920403 920403 920403 Stacked 920415 920415 920415 920415 920416 920416 920416 920416 event of F1 12:48:12.24 21:38:26.52 03:06:07.73 07:32:06.91 08:33:32.32 10:18:15.11 12:19:31.99 13:58:01.08 event of F3 02:04:39.68 02:16:16.94 02:45:14.81 03:06:16.65 02:51:15.71 05:26:48.87 17:20:02.06 17:28:34.93 37.7583 37.7615 37.7568 37.7583 37.7568 37.7570 37.7567 37.7572 37.7587 37.7577 37.7590 37.7580 37.7555 37.7580 37.7577 37.7587 37.7587 37.7582 15.0027 15.0037 15.0088 15.0038 15.0080 15.0078 15.0093 15.0057 15.0047 15.0052 15.0018 15.0035 15.0047 15.0045 15.0045 15.0018 15.0045 15.0040 Depth −2.02 −1.66 −1.79 −2.09 −1.82 −1.81 −1.75 −1.84 −2.10 −2.06 −2.29 −2.20 −2.57 −2.12 −2.12 −2.27 −2.03 −2.15 No RMS 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 0.02 0.02 0.01 0.04 0.01 0.02 0.01 0.02 0.01 0.03 0.03 0.02 0.03 0.02 0.02 0.02 0.03 0.03 No - number of P arrivals; RMS - travel-time residual root mean square (s) The “stacked” events (#1 and #10) have been used as master events; the two sets of events were analysed separately and finally all together (with #10 as master event) since they have fairly similar waveforms (Table 4); the relative timing (dt) for pairs of earthquakes has been obtained on a 2.05 s window of signal (256 points) starting about 0.25 s before the first arrival, thereby allowing the complete sampling of the P wave train evidenced by the polarization analyses performed by Falsaperla et al. (2002). The quality of comparisons between the events at different stations was better than Q = 90 and better than Q = 95 between most events (see Table 4); the relative timing between the dt shows errors less than ± 0.0038 s (mean error between 0.9 and 2.9 ms; Table 4). For each family, we performed the relative locations using the parameters (take-off angles and azimuths) obtained by analytical locations of master events and a velocity of 3.0 kms−1 in the source region (Hirn et al., 1991). The location differences between events ranges between several meters to less than two hundred meters and associated errors range from 6.5 and 15.5 m (Table 4). The relocated seismicity, in Fig. 4, clearly describes the geometry of a vertically (210 m high) elongated volume covering a restricted horizontal area of ca. 45 × 50 m. 6. DISCUSSION AND CONCLUSIONS In this paper, a high precision relocation of LP multiplets recorded on Mt. Etna has been reported for the first time. This seismicity, occurring in April and September 1992, 670 Stud. Geophys. Geod., 50 (2006) High Precision Locations of LP Events on Mt. Etna: Reconstruction of the Fluid-Filled Volume Fig. 4. Epicentral map of located and relocated events. The insets represent the enlarged map, NS and EW cross-sections of the relocated events. Origin coordinates and depth are referred to #10 event. characterized the summital area of Mt. Etna close to the Northeast crater. This crater, created in 1911, has been successively modified by numerous explosive and effusive episodes and seems to have evolved as an independent summit event (La Delfa et al., 2003). Multiplets are due to the repeated occurrence of almost identical source mechanisms and their relative relocation allows obtaining the spatial distribution of the events with high precision; unlike tectonic events, the relocation of LPs does not define fault planes but rather volumes (Battaglia et al., 2003). The results of relocating 1992 Mt. Etna LP multiplets define a vertically elongated volume, about 50 m long, 45 wide and 210 m high; the uncertainties in the relocations are between 6.5 and 15.5 m. The occurrence of LPs is attributed to the resonance of a fluid filled volume in response to a trigger. The geometries proposed by different authors at different volcanoes vary in shape for the resonant structure: spherical for Kazahaya et al. (1994), tabular such as a crack or a fissure for Aki et al. (1977) and Chouet (1988), and cylindrical for Chouet (1985). The oscillation can be induced by various mechanisms such as a mechanical trigger (Chouet, 1988, 1992), a flow impediment (White et al., unpublished data) and changes in channel geometry (Julian, 1994); Falsaperla et al. (2002) concluded that the 1992 LP events were originated by collapses of the NEC floor that triggered the excitation of the magma feeding system located beneath Northeast Crater. They applied the fluid-filled crack model (Chouet, 1988, 1992) obtaining a crack of 400 × 400 m, but they also concluded that LP events may originate in a segment of the magma conduit beneath NEC. Stud. Geophys. Geod., 50 (2006) 671 S. Gambino Table 4. Event relocation results. Family n master n event F1 F1 F1 F1 F1 F1 F1 F1 F3 F3 F3 F3 F3 F3 F3 F3 F3 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 02 03 04 05 06 07 08 09 11 12 13 14 15 16 17 18 10 An. St. Mean Qw 5 6 6 6 5 6 6 6 6 7 7 6 6 7 7 7 6 T. err. [msec] 0.903 0.958 0.965 0.979 0.959 0.917 0.957 0.944 0.958 0.964 0.974 0.950 0.970 0.979 0.980 0.973 0.963 1.5 1.3 0.9 1.4 1.3 1.6 0.9 1.5 2.8 2.8 2.1 3.2 2.7 2.1 1.7 2.2 1.0 Dx [m] −22.8 3.3 −23.2 −4.8 17.9 −46.9 −21.3 9.5 −7.5 −27.1 −4.2 −22.4 −25.2 −6.3 −17.2 −12 −7.5 Dy [m] 30.1 4.9 16.7 22.1 44.6 30.8 9.3 22 4.6 8.93 0.93 8.08 41.6 21.4 21.9 28.7 4.6 Dz [m] −154 35.3 −31 6.2 −5.9 32.9 −37.2 56.8 9.1 6.4 26.7 14.2 −63.2 −51 −75.2 −95.1 9.1 Ril. err. [m] 14.0 7.4 6.5 14.8 6.8 8.0 6.6 8.5 12.8 11.1 9.1 15.5 11.7 9.9 8.7 9.5 6.8 An. St. - number of stations used in relocation; Mean Qw - mean of the quality factor at different stations; T. err. - mean relative time error; Dx, Dy, Dz - location differences and associated error (Ril. err.) with respect to the master event. The results of the high precision relocation show that the geometry of the Mt. Etna 1992 resonator is a cylindrical fluid filled conduct rather than a crack. 3D absolute locations have enabled improving source location of LP events, indicating that the conduit segment belongs to the NEC system and is quite shallow (ca. 2000 m a.s.l.). In conclusion, LP events represent, within the mechanisms and evolution of the 1991-93 eruption, a minor and local phenomenon confined to the upper NEC system that probably originated from floor collapses inside a partially filled conduit. In addition, the obtained conduit geometry is in accordance with Bonaccorso and Davis (1999) that, modelling ground deformation related to the violent NEC explosive activity on December 23, 1995, considered a vertical conduct with radius of 20 m to justify the variations measured on tilt signals. Finally, this paper represents an example of the potential of the LP high precision relocation technique in identifying important volcanological parameters such as the geometry of a magma-filled conduit, crack or chamber. References Aki K., Fehler M. and Das S., 1977. Source mechanism of volcanic tremor: Fluid driven crack models and their application to the 1963 Kilauea eruption. J. Volcanol. Geotherm. Res., 2, 259−287. 672 Stud. Geophys. Geod., 50 (2006) High Precision Locations of LP Events on Mt. Etna: Reconstruction of the Fluid-Filled Volume Aki K. and Richards P.G., 1980. Quantitative Seismology, Theory and Methods. W H Freeman, San Francisco. Aloisi M., Cocina O., Neri G., Orecchio B. and Privitera E., 2002. Seismic tomography of the crust underneath the Etna volcano, Sicily. Phys. Earth Planet. Inter., 134, 139−155. Alparone S. and Gambino S., 2003. High precision locations of multiplets on south-eastern flank of Mt. Etna (Italy): reconstruction of fault plan geometry. Phys. Earth Planet. Inter., 135, 281−289. Barberi F., Carapezza M.L., Valenza M. and Villari L., 1993. The control of lava flow during 1991−1992 eruption of Mount Etna. J. Volcanol. Geotherm. Res., 56, 1−34. Battaglia J., Got J.L. and Okubo P., 2003. Location of long-period events below Kilauea Volcano using seismic amplitudes and accurate relative relocation. J. Geophys. Res., 108, 2553, doi:101029/2003JB002517. Battaglia J., Thurber C.L., Got J.L., Rowe C.A. and White R.A., 2004. Precise relocation of earthquakes following the June 15, 1991 explosion of Mount Pinatubo (Philippines). J. Geophys. Res., 109, B07302, doi:10.1029/2003JB002883. Baher S., Thurber C., Roberts K. and Rowe C., 2003. Relocation of seismicity preceding the 1984 eruption of Mauna Loa Volcano, Hawaii: Delineation of a possible failed rift. J. Volcanol. Geotherm. Res., 128, 327−339. Bonaccorso A. and Davis P.M., 1999. Models of round deformation from vertical volcanic conduicts with application to eruptions of Mount St. Helens and Mount Etna. J. Geophys. Res., 104, 10531−10542. Brancato A. and Gresta S., 2003. High precision relocation of microearthquakes at Mt. Etna (1991−1993 eruption onset): a tool for better understanding the volcano seismicity. J. Volcanol. Geotherm. Res., 124, 219−239. Calvari S., Coltelli M., Pompilio M. and Neri M., 1994. 1991−1993 Etna eruption: geological observation and chronology of eruptive events. Acta Vulcanol., 4, 1−14. Cattaneo M., Augliera P., Spallarossa D. and Eva C., 1997. Reconstruction of seismogenetic structures by multiplet analysis: an example of Western Liguria, Italy. Bull. Seismol. Soc. Amer., 87, 971−986. Chouet B.A., 1985. Excitation of buried magmatic pipe: a seismic source model for volcanic tremor. J. Geophys. Res., 90, 1881−1893. Chouet B.A., 1988. Resonance of a fluid-driven crack: radiation properties and implications for the source of long-period events and harmonic tremor. J. Geophys. Res., 93, 4375−4400. Chouet B.A., 1992. A seismic model for the source of long period events and harmonic tremor. In: P. Gasparini, R. Scarpa and K. Aki (Eds,), Volcanic Seismology, IAVCEI Proc. in Volcanology, 3, Springer-Verlag, Berlin, pp. 133−156. Chouet B.A., Dawson P.B., Falsaperla S. and Privitera E., 1994. A characterization of long-period events recorded during the eruptive activity of Mount Etna, Italy, in 1992. Acta Vulcanol., 4, 81−86. Chouet B.A., 1996. Long-period volcano seismicity: its source and use in eruption forecasting. Nature, 300, 309−316. Coltelli M., Pompilio M., Del Carlo P., Calvari S., Pannucci S. and Scribano V., 1998. Etna: eruptive activity. Acta Vulcanol., 10, 141−149. Falsaperla S., Privitera E., Chouet B. and Dawson P., 2002. Analysis of long period events recorded at Mount Etna (Italy) in 1992 and their relationship to eruptive activity. J. Volcanol. Geotherm. Res., 114, 419−440. Stud. Geophys. Geod., 50 (2006) 673 S. Gambino Ferrucci F. and Patanè D., 1993. Seismic activity accompanying the outbreak of the 1991−1993 eruption of Mount Etna (Italy). J. Volcanol. Geotherm. Res., 57, 125−135. Frèmont M.J. and Malone S.D., 1987. High precision relative locations of earthquake at Mount St. Helens, Washington. J. Geophys. Res., 92, 10223−10236. Got J.L., Fréchet J. and Klein F.W., 1994. Deep fault plane geometry inferred from multiplet relative relocation beneath the south flank of Kilauea. J. Geophys. Res., 99, 15375−15386. Haase J.S., Shearer P.M. and Aster R.C., 1995. Constraints on temporal variations in velocity near Anza, California, from analysis of similar event pairs. Bull. Seismol. Soc. Amer., 85, 194−206. Hirn A., Nercessian A., Sapin M., Ferrucci F. and Wittlinger G., 1991. Seismic heterogeneity of Mt. Etna: structure and activity. Geophys. J. Int., 105, 139−153. Jones J.P., Thurber C.H. and Lutter W.J., 2001. High precision location of pre eruption seismicity at Mount Pinatubo, Philippines, 30 May−3 June, 1991. Phys. Earth Planet. Inter., 123, 221−232. Julian B.R., 1994. Volcanic tremor: nonlinear excitation by fluid flow. J. Geophys. Res., 99, 11859−11877. Kazahaya K., Shinobara H. and Saito G., 1994. Excessive degassing of Izu-Oshima volcano: magma convection in a conduict. Bull. Volcanol., 56, 207−216. La Delfa S., Innocente V., Patanè G. and Tanguy J.C., 2003. Correlation between local stress field and summit eruption of Mount Etna: the 27 March 1998 event. Geol. Carpath., 54, 251−260. Lahr J.C., 1989. Hypoellipse Version 20*: a Computer Program for Determining Local Earthquake Hypocentral Parameters, Magnitude, and First Motion Pattern. United States Department of the Survey, Menlo Park, California, Open File Report 89-116. Lo Giudice E., Patanè G., Rasà R. and Romano R., 1982. The structural framework of Mt. Etna. Mem. Soc. Geol. Ital., 23, 125−158. Monaco C., Tapponnier P., Tortorici L. and Gillot P.Y., 1997. Late Quaternary slip rates on the Acireale-Piedimonte normal faults and tectonic origin of Mt. Etna (Sicily). Earth Planet. Sci. Lett., 147, 125−139. Moriya H., Niitsuma H. and Baria R., 2003. Multiplet-clustering analysis reveals structural details within the seismic cloud at the Soultz geothermal field, France. Bull. Seismol. Soc. Amer., 93, 1606−1620. Patanè D., Cocina O., Falsaperla S., Privitera E. and Spampinato S., 2004. Mt. Etna volcano: a seismological framework. In: S. Calvari, A. Bonaccorso, M. Coltelli, C. Del Negro and S. Falsaperla (Eds.), The Mt. Etna Volcano. AGU, Washington, D.C., 147−165. Patanè D., Chiarabba C., De Gori P. and Bonaccorso A., 2003. Magma ascent and the pressurization of Mt. Etna’s volcanic system. Science, 299, 2061−2063. Pechmann J.C. and Kanamori H., 1982. Waveform and spectra of preshocks and aftershocks of the 1979 Imperial Valley, California, earthquake: evidence for fault heterogeneity? J. Geophys. Res., 87, 10579−10579. Poupinet G., Ellsworth W.L. and Fréchet J., 1984. Monitoring velocity variation in the crust using earthquake doublets: an application to the Calaveras fault, California. J. Geophys. Res., 89, 5719−5731. Rowe C.A., Thurber C.H. and White R.A., 2004. Relocation of volcanic event swarms at Soufriere Hills volcano, Montserrat, 1995−1996. J. Volcanol. Geotherm. Res., 134, 199−221. Thurber C.H., 1993. Local earthquake tomography: velocity and VP/VS-theory. In: H.M. Iyer and K. Hirahara (Eds.), Seismic Tomography: Theory and Practice. Chapman and Hall, London, 563−583. Wolfe C.J., Okubo P.G. and Shearer P.M., 2003. Mantle fault zone beneath Kilauea Volcano, Hawaii. Science, 300, 478−480. 674 Stud. Geophys. Geod., 50 (2006)