The Mt. Etna January 9, 2001 Seismic Swarm

Open Geosci. 2016; 8:514–522

Research Article Open Access

Salvatore Gambino* Tilt offset associated with local seismicity: the Mt. Etna January 9, 2001 seismic swarm.

DOI 10.1515/geo-2016-0045 bubble sensors; long-base devices have two orthogonal Received December 23, 2015; accepted June 12, 2016 fluid-filled tubes positioned inside tunnels where fluid levels are measured at the ends of the tubes. Tiltmeters are Abstract: On the 9th of January 2001 a seismic swarm on a powerful tool for volcano monitoring and tilt changes the southeastern flank of Mt. Etna at 3.5 km beside sea have accompanied the rapid rise of magma and formation level (b.s.l.), caused co-seismic variations on short and of dikes and eruptive fissures at Mt. Etna (e.g. [4]). long baseline tiltmeters of the Mt. Etna permanent tilt net- Faulting causes tilt changes with amplitudes related work. to source-station distance and earthquake magnitude [5]; Taking account of the geometry and mechanism of the ac- they are generally characterized by impulsive variations tive tectonic structure obtained by seismological studies, with simultaneous recordings at different stations in cor- the theoretical tilt linked to the faulting source was calcu- respondence with the earthquake. Changes are generally lated at multiple different recording stations. It was found small for shallow (0–5 km depth) local earthquakes at that the amount of measured deformation exceeded that Mt. Etna, comprised between 0.1 and 1.5 microradians [3]. which was generated seismically, indicating that much of These permanent offsets are assumed to represent the the deformation along the fault was aseismic. change produced by the fault dislocation in a tectonic con- The 9 January 2001 episode represents a shear response to text (e.g. [5, 6]) while in an active volcanic setting, the ten- a local stress caused by a volcanic source that acted in the sile component (e.g. a dike) can dominate the deforma- period preceding the 2001 eruption. Tilt data also suggest tion. Analyses of co-seismic variations in different areas a marked slip of 70-140 cm along the fault, probably due to have highlighted how continuous borehole tilt data may the presence of fluids. also comprise seismic shaking effects, such as movements Keywords: Seismic swarm; coseismic tilt; fault plane ge- on cracks and fractures near the instrument site, insta- ometry; aseismic slip bility of instruments or translational ground acceleration caused by passing waves [1, 5, 7, 8] that cause larger variations [9–11]. Nonetheless, these effects are not present on 1 Introduction a long-base device [1, 12, 13]. Tilt changes associated with a seismic swarm (M = 3.5) occurring in the eastern flank of Mt. Etna Fault reactivation gives rise to earthquakes and ground max on January 9, 2001 (Fig. 1) were examined. This swarm surface deformation caused by sudden slippage across was characterized by tens of events with similar wave- the fault surface. Investigations into ground deformation forms whose high precise relocation, with focal planes, are important to improve our knowledge of tectonic pro- allowed accurate reconstruction of the seismogenic fault cesses. Continuous high precision measurements (like tilt, geometry [14]. strain) are considered the most appropriate to investigate Amplitudes and azimuths of the coseismic tilt offsets the small extent of ground deformation related to fault are compared with those from the dislocation theory using rupturing [1]. parameters obtained by seismology and possible reasons On Mt. Etna, a tilt network measures ground angular for their discrepancy are discussed. movements using borehole and long-base instruments [2, 3]. Borehole instruments use high precision electrolytic 2 Mt. Etna *Corresponding Author: Salvatore Gambino: Istituto Nazionale di Geofisica e Vulcanologia- Osservatorio Etneo, Piazza Roma 2, Located between the compressive domain of Western- 95123 Catania, Italy, E-mail: [email protected]; tel.: 39-95-7165877, Central Sicily and the tensional domain of the Calabrian fax: 39-957165826

Arc, Mt. Etna (Fig. 1) has formed at the intersection of two Genesis of this seismicity may have three main regional fault systems, which have NNW-SSE and NE-SW sources: regional tectonic stress, local stress induced trends (e.g. [15, 16]), respectively (Fig. 1). by dike propagation or stress due to slower inflating sources [23, 25, 26]. In particular, seismic activity in the southeastern sector of volcano, located between 3 and 8 km b.s.l., has been associated with an E-W compression induced by a pressurizing source just westwards and at the same depth [22]. Following the effusive flank eruptions of December 1991- March 1993 there was a recharge phase accompanied by near constant and marked inflation of the overall volcano that lasted until 2001. This phase culminated with the July - August 2001 and the October 2002 – January 2003 effusive-explosive flank eruptions [27]. The 2001 eruption is one of the most studied events both from volcanological and geophysical viewpoint. It occurred from seven eruptive fissures, showed high levels of explosivity and produced 5–10 × 106 m3 of ash and ~25 × 106 m3 of lava [28]. The final accelerated intrusive process of the 2001 eruption and the opening of the eruptive fractures was accompanied by important seismic activity and marked ground deformation was recorded on the 12–17 of July [29, 30]. In particular, between late 2000 and July 2001, the Figure 1: Map of permanent tilt networks and surface faults of Mt. recharge was accompanied by seismic activity, including Etna. Lower right inset map shows the main regional fault systems: an intense seismic swarm on January 9, 2011, that spread MF=Messina-Fiumefreddo line, ME=Malta Escarpment. Dashed lines across a wide area mainly in the southern and eastern part define the sliding sector. The dashed box shows the area drawnin Fig. 2. of the volcano [23].

The interaction between regional stress, dike-induced rifting and gravity causes a fairly continuous and roughly 3 Tilt network eastward and downward motion of Mt. Etna’s eastern flank (e.g. [17]). This sliding area (Fig. 1) is delimited to the north The Istituto Nazionale di Geofisica e Vulcanologia – Osser- by the Pernicana–Provenzana Fault System (e.g. [18, 19]) vatorio Etneo began using bubble sensor tiltmeters on Mt. and to the south by the Trecastagni and Tremestieri faults Etna in 1977 [2]. The Mt. Etna permanent tilt networks cur- (e.g. [20, 21]) and the décollement surface is located at a rently comprise 14 bi-axial instruments installed in shal- depth of 3 km b.s.l. ([22] and references therein). low boreholes and one long-base instrument positioned Earthquakes at Mt. Etna, usually defined as seismo- inside two 80-m-long tunnels at the Volcanological Obser- tectonic events, can generally be divided in two ma- vatory of Pizzi Deneri [3]. jor groups: volcanic-tectonic (VT) earthquakes are gener- Borehole instruments use a high precision electrolytic ated by tectonic stress and/or by stress arising from ris- bubble sensor [31] to measure the angular movement ∂ ∂ x ing magma (e.g. [23]) and long-period events most likely ( z/ x) and each tiltmeter has two perpendicular axes, y x driven by stress changes caused by an intermittent de- and , with the -axis generally oriented toward the vol- gassing process occurring at depth [24]. cano summit. This sensor orientation is usually chosen to VT earthquakes occur mainly in the form of swarms better detect ground tilt associated with pressure change composed of micro or small earthquakes which seldom in an isotropic pressure source located under the active exceed magnitude 4. The deep (ca. 15–30 km) and inter- crater (i.e., producing a symmetrical ground deformation mediate (ca. 5–15 km) have high-frequency seismic waves field). (mostly above 4–5 Hz), the events <5 km show medium to In 2001, the tilt network configuration, as reported in high frequencies [25]. Fig. 1, included eight active bi-axial instruments installed 516 Ë Salvatore Gambino in shallow boreholes at about 3 m depth and one long In [14] the authors re-located a group of multiplets us- baseline instrument. Four stations were equipped with ing a cross-spectrum method [36] and obtained accurate AGI (Applied Geomechanics) 510 tiltmeters with a preci- relative relocations with uncertainty of few tens of me- sion of 0.01 microrad (DAM, MDZ, MEG, MMT), three with ters. The relocated events clearly describe the geometry AGI 722 model with a precision of 0.1 microrad (CDV, MNR, of the seismogenic structure; the events lie on a 77° dip- MSC), while a Kinemetrics instrument with a precision of ping plane, oriented ENE-WSW (N55°E) in line with one 0.025 microrad was set up at SPC station. The real accu- of the planes of the focal mechanisms (Fig. 2) for the four racy is however affected by environmental noise linked to stronger events (2.8≤Md≤3.5). Fault plane solutions (FPSs) temperature, rainfall or air pressure. At a depth of a few at Mt.Etna are commonly performed under the basic as- meters, the thermoelastic effects may produce a noise both sumption of a double couple source. In a volcano the in seasonal and daily variations (e.g. [1, 32]) so that the in- possibility of having tensile crack, corrige compensated- strumental accuracy is not less than 0.1 microradians. linear-vector-dipole (CLVD) and isotropic source mecha- In 1997, a long-baseline instrument comprising two nisms is high in particular for events occurring during dike 80 m long orthogonal tubes filled with mercury, whose intrusion and fracture propagations; In particular, [37, 38] levels are measured at ends of the tubes, was positioned found evidence for CLVD and isotropic source mechanisms along the tunnels at the Volcanological Observatory of during 1991-93 and 2001 Mt. Etna episodes. Pizzi Deneri (PDN, 2850 m a.s.l.) [2]. This device has a low However, the 9 January 2001 events are several kilo- temperature noise and an accuracy of ca. 0.05 microradi- meters away from the eruptive centers, and at 3–5 km ans. depth with high frequency (5-15 Hz) waveforms. For these Data acquisition varies from 1 sample every minutes features the basic assumption of a double couple source to 1 sample every 30 minutes, including acquisition of tilt seems to be reasonable [14, 39]. components, air and ground temperatures, air pressure and instrumental control parameters. In recent years the network has been improved with deep stations using ad- vanced high resolution (<5 nanoradians) self-levelling instruments [33].

4 Seismicity

The seismic swarm analyzed here started on January 9, 2001 at 02:51 GMT with around one hundred VT earthquakes with M≥1.0 that affected the southeastern flank of Mt Etna. This seismicity clearly showed P and S phases with evident onsets and predominant frequencies in the range of 5-15 Hz and very similar waveforms (multiplets) [14]. The most energetic event (Md=3.5) occurred at 02:51 Figure 2: Location and E-W cross-section of the 9, January 2001 and was preceded by 3 other events with magnitudes of seismic swarm. Focal mechanisms solutions for the four stronger 2.5, 2.7 and 2.8 occurring in the previous 45 seconds (tab. events (2.8≤Md≤3.5) are reported. Three focal solutions show pure strike-slip fault planes. Numbers are referred to in Table 1. 1). The aftershock “sequence” comprised a magnitude 3.0 shock at 03:32 and one of 3.1 at 04:31 (Table 1). The epicen- ters were within an area of about 2 km2 near the town of Zafferana Etnea [34]; focal depths ranged between 3and 5 km (Fig. 2). Fig. 3 reports the earthquake minute rate and related strain release from 02:00 to 05:00 GMT. The strain 5 Tilt changes release was obtained as the square root of the seismic energy computed by using the Richter (1958) relationship: Impulsive coseismic tilt variation (offset) events were detected by three tiltmeters. The signals recorded at PDN log E(erg) = 9.9 + 1.9M − 0.024M2 (1) show more precise data since they had a higher sampling Tilt offset associated with local seismicity Ë 517

Table 1: Location parameters of the analyzed earthquakes (extracted by [14]).

N Date Origin time Md Latitude Longitude Depth Gap No RMS ERZ ERH 1 09/01/01 02:51:14.98 2.5 37.705 15.070 3.92 61 18 0.11 0.5 0.3 2 09/01/01 02:51:33.70 2.7 37.701 15.071 3.78 55 23 0.10 0.4 0.3 3 09/01/01 02:51:51.09 2.8 37.702 15.071 3.59 54 24 0.11 0.4 0.3 4 09/01/01 02:51:58:44 3.5 37.703 15.076 3.77 57 36 0.14 0.2 0.2 5 09/01/01 02:56:01.59 2.1 37.704 15.064 4.52 73 17 0.11 0.5 0.5 6 09/01/01 03:04:14.32 1.8 37.704 15.068 5.13 98 12 0.06 1.0 0.8 7 09/01/01 03:23:51.57 1.8 37.704 15.065 4.36 91 15 0.08 0.4 0.7 8 09/01/01 03:32:18.56 3.0 37.708 15.080 3.70 54 35 0.16 0.2 0.2 9 09/01/01 03:53:19.32 1.8 37.705 15.068 4.49 96 13 0.04 0.7 0.7 10 09/01/01 04:31:39.45 3.1 37.708 15.078 3.78 54 36 0.16 0.2 0.2 11 09/01/01 04:39:03.01 2.1 37.706 15.073 4.72 74 19 0.07 0.6 0.6 12 09/01/01 08:13:17.75 2.5 37.702 15.071 3.59 54 22 0.05 0.4 0.3 13 09/01/01 16:28:53.01 2.0 37.702 15.072 4.63 54 21 0.10 0.5 0.4 14 09/01/01 21:20:33.36 1.9 37.701 15.069 3.55 91 18 0.08 0.4 0.4 15 11/01/01 14:01:51.05 2.8 37.702 15.074 3.63 54 42 0.14 0.2 0.2 16 25/01/01 06:32:12.87 2.3 37.701 15.069 3.72 53 27 0.08 0.4 0.3 17 04/02/01 10:25:20.11 1.7 37.700 15.064 3.42 87 10 0.10 0.7 0.6 18 04/02/01 10:25:24.58 2.2 37.700 15.066 3.85 45 24 0.14 0.5 0.4 19 25/03/01 18:25:30.99 1.7 37.699 15.066 3.51 77 13 0.06 0.7 1.0 Md = duration magnitude; GAP = azimuthal gap (degrees); RMS = travel-time residual root mean square (s); No = number of P and S arrivals. rate (1 minute) and greater sensitivity. These signals high- ca. 77° dip, N55°E oriented plane located 3 km north-west light that the coseismic tilt (Figure 3C) was linked to the of Zafferana Etnea (Fig. 1). main local earthquake at 02:51 (together with the events The cumulative seismic moment release associated occurring in the first 2 minutes), and had a sudden varia- with the seismic events of the 02:51-03:00 seismic events tion of ca. 0.4 microradians on the radial component and was estimated by using the formula [40] for Mt. Etna earth- 0.2 microradians on the tangential one. Tilting at PDN then quakes: continued for about 38 minutes, cumulating respectively M M 0.20 and 0.18 microradians on the two components. log( o) = (17.60 ± 0.37) + (1.12 ± 0.10) * L (2) Between 2:30 and 3:00 GMT, the CDV station showed where ML is the local Magnitude of each event. After the a change in the tilt vector modulus of 0.5 microradians conversion of Md in ML by using the [41] relation: (Fig. 4), while a large variation of more than 4.0 microradians was recorded at the SPC station (Fig. 4). It is notewor- ML = 1.164(±0.011) * Md − 0.337(±0.020) (3) thy that between 3:00 and 3:30 GMT, the tangential component at the CDV and SPC stations (Fig. 4) showed a pos- A Mo no greater than = 4.1*1022 dyne-cm was ob- itive change of 0.4-0.5 microradians. Tilt vectors related to tained. In light of this result, assuming a medium rigidity 02:30-03-00 and 03:00-03:30 are reported in Fig. 5 and Ta- (µ) of a shear modulus of 10 GPa [42], a 2 km long and 1 km ble 2. No variations were recorded at the more distant sta- wide fault (S = 2 km2) and using the general relation of [43]: tions on the western (MMT, MSC, MMT) and northern (DAM and MNR) flanks. MDZ station was out of order. Mo = µ * S * u (4) A displacement (u) of 21 cm was obtained. Considering fault geometry parameters and 21 cm 6 Tilt expected from seismicity of right strike-slip, I calculated the expected tilt at all the network stations (Table 2) for a tabular dislocation For the 9 January 2001 swarm, Alparone and Gambino model striking N55°E, dip 77°, located at the earthquake (2003) determined a right-lateral strike mechanism along hypocenter by using the [44] dislocation model. 518 Ë Salvatore Gambino

Table 2: Recorded and calculated tilt at different stations of the network. Values in brackets are referred to 02:54-03:30 for PDN and03:00- 03:30 for CDV and SPC time periods. Modules are in microradians.

RECORDED CALCULATED Station module direction Module Direction PDN 0.44 (0.27) N253°E (N274°E) 0.068 N238°E CDV 0.5 (0.56) N277°E (N210°E) 0.155 N285°E SPC 4.20 (0.4) N215°E (N75°E) 0.180 N10°E MSC 0.0 0.014 MGT 0.0 0.011 MMT 0.0 0.007 MNR 0.0 0.014 DAM 0.0 0.010

Figure 3: Earthquakes rate (occurrence every minute) (a) and related strain release (b). Coseismic variations detected at PDN long-base station (c). Figure 4: Coseismic tilt variations detected at CDV and SPC stations. Tilt offset associated with local seismicity Ë 519

addition to the real ground tilt, larger variations for different effects related to seismic shaking. The recorded data at SPC during the January 9th events is an example of this; the large variation of tilt (>4.0 microradians) not observed at CDV (2.5 km away) is probably linked to seismic shaking effects (e.g. [10]). The short transients in the CDV tilt data instead suggest that it was due to real ground tilt and not instrument response to seismic shaking [6]. No shaking effects involve a long-base device such as the mercury instrument of Pizzi Deneri Observa- tory [13]. The January 9, 2001 seismic swarm started at 02:51 GMT and 6 VT events (Fig. 3A) were recorded in two minutes with an M=3.5 main event. Fig. 5 shows that the direction of the tilt vector as measured by the long base tiltmeter and CDV match well the direction obtained by using the [44] model between 2:30 and 3:00 GMT. How- ever, there is a magnitude difference considering that the calculated value is roughly 30% less than the recorded value at CDV and ca. 15% less than that at PDN. This means that most likely only a part of the stick-slip obtained by modeling is related to the co-seismic effects, along the 9 January 2001 active fault, suggesting that most of the slip over the fault must be aseismic. This observation is common in volcanic areas (e.g. [48, 49]) and also

Figure 5: Comparison between observed (blue) and modeled (red on Mt. Etna. In particular [19] concluded that only 30% of and purple) tilt vectors. The numbers are referred to 02:30-03-00 the total deformation of the Pernicana Fault is attributable GMT (1) and 03:00-onwards GMT (2) time periods. Values less than to co-seismic displacements and [50] showed a seismic ef- 0.1 microrads (the bore-hole sensitivity) are not reported. ficiency lower than 30% for the Fiandaca Fault (see Fig. 1 for their locations). If this is the case for the 9 January 2001 swarm, only The tilt vector at the PDN station was 0.068 microradi- the 15-30% of the total deformation is attributable to co- ans, N238°E oriented, and 0.155 microradians, N285°E for seismic displacements. Hence, the active fault has been the CDV station (Table 2). affected by a larger slip (70–140 cm) with respect tothe Recorded and calculated values have been reported on 21 centimeters suggested by seismicity. a map as tilt vectors in Fig. 5. MGT, MNR, DAM, MSC, and MMT tilt stations are lo- Moreover, I applied the same procedure to the 3:00- cated at some distance from the earthquake foci (Fig. 1) 5:00 period cumulative seismic moment, obtaining a dis- and did not detect variations. Using seismological param- placement of 8 cm and very small tilt vectors (oriented as eters (Tab. 2) at these stations, the module predictions previous) of 0.026 and 0.050 microradians respectively at were low (7 to 14 nanorads). It is interesting to note that PDN and CDV. if I consider a fault displacement 3.3-6.6 times larger than that obtained by seismology (as suggested by recorded data) the expected tilt is less than instrumental sensitiv- 7 Discussion and conclusions ity (0.1 microradians) in all cases and I should not observe any change even though fault displacement occurred. A residual tilt offset associated with seismic events isas- In summary the seismic source of the 9 January 2001 sumed to represent the change produced by the fault dis- event was not close to Mt. Etna’s internal magma sources location [5]. Analyses of co-seismic variations on bore-hole (e.g. [51, 52])., it was more deeper than the detachment sur- instruments (e.g. [11, 45–47]) have highlighted how contin- face of the eastern flank [22] and seems to have tectonic uous tilt measurements at the short-baseline may have, in features rather than volcanic. An analysis of tilt data help to exclude the presence of magma. 520 Ë Salvatore Gambino

Initially, I ran tensile models using the same fault ge- [2] Bonaccorso A., Campisi O., Falzone G., Gambino S., Continu- ometry, with different ranges of reasonable openings in or- ous tilt monitoring: lesson from 20 years’ experience at Mt. der to obtain the deformations linked to a possible dike Etna. In: A. Bonaccorso, S. Calvari, M. Coltelli, C. Del Negro, and S. Falsaperla (eds.), Mt. Etna: Volcano Laboratory, Ameri- intrusion along the fault. Table 3 reports the expected tilt can Geophysical Union, Washington D.C., 2004, 307-320, DOI: signals considering an opening of 10 cm and Fig. 5 shows 10.1029/143GM19. the obtained tilt vectors. Vectors are quite different to those [3] Gambino S., Falzone G., Ferro A., Laudani G., Volcanic pro- recorded and seem to exclude an intrusive process. How- cesses detected by tiltmeters: a review of experience on Sicilian ever, the presence of fluids in this sector of the volcano volcanoes, J. Volcanol. Geotherm. Res. 2014, 271, 43-54, 2014. [4] Aloisi M., Bonaccorso A., Gambino S., Imaging compositive dike is possible; recent magnetotelluric surveys (8 km depth) propagation (Etna, 2002 case). J. Geophys. Res., 2006, 111, on the eastern flank of Mt. Etna have shown several high- B06404, doi: 10.1029/2005JB003908. conductivity zones, suggesting a diffuse presence of hy- [5] McHugh S., Johnston M.J.S., An analysis of coseismic tilt drothermal activity and fluid circulation [53]. Moreover, changes from an array in Central California, J. Geophys. Res., seismicity is characterized by multiplets that can arguably 1977, 82, 5692-5698. be interpreted as caused by a single persistent, and pos- [6] Anderson K., Lisowski M., Segall P., Cyclic ground tilt associated with the 2004–2008 eruption of Mount St. sibly non-destructive, source; this factor simply does not Helens. J. of Geophys. Res., 2010, 115 (B11), B11201, agree with a purely tectonic origin of these earthquakes. doi:10.1029/2009JB007102. [7] Mellors, R., Chatelain J.L., Isacks B.L., Hade G., Bevis M., Prevot, M., A tilt and seismicity episode in the New Hebrides (Vanuatu) Table 3: Calculated modulus and direction of tilt vectors for a tensile Island Arc, J. Geophys. Res., 1991, 96, 16535-16546. model with opening of 10 cm along a 77°dip, N55°E oriented plane [8] Fournier N., Jolly A.D., Miller C., Ghost tilt signal during transient located 3 km north-west of Zafferana Etnea. Values obtained at the ground surface deformation events: Insights from the Septem- other stations are <0.01 microdarians. ber 3, 2010 Mw7.1 Darfield earthquake, New Zealand, Geophys. Res. Lett., 2011, 38, L16305, doi:10.1029/ 2011GL048136. Modulus Direction [9] Bonaccorso A., Gambino S., Impulsive tilt variations on Mount PDN 0.054 N150°E Etna (1990-93), Tectonophysics, 1997, 270, 1-2, 115-125. CDV 0.354 N341°E [10] Gambino S., Coseismic and aseismic tilt variations on Mount SPC 0.150 N351°E Etna. Pure Appl. Geophys. 2002, 159 11, 2751-2762. [11] Gambino S., Campisi O., Falzone G., Ferro A., Guglielmino F., Laudani G., Saraceno B. Tilt measurements at Vulcano Island. th Annals of Geophysics, 2007, 50, 233-247. In conclusion, the 9 of January swarm is the result of [12] Bonaccorso A., Patane’ D., Shear response to an intrusive a marked faulting episode along a pre-existing structure of episode at Mt. Etna volcano (January 1998) inferred through the eastern flank that may have been aided by the presence seismic and tilt data, Tectonophysics, 2001, 334, 61-75. of fluids. A local stress, induced by the recharge that pre- [13] Bonforte A., Gambino S., Guglielmino F., Obrizzo F., Palano M., ceded the Mt. Etna 2001 eruption [22], is the most reliable Puglisi G., Ground deformation modelling of flank dynamics prior to the 2002 eruption of Mt. Etna. Bull. Volc., 2007, 69, 757- source of the dislocation. The variations recorded between 768. 03:00 and 3:30 at the three tilt stations (Fig. 5) differ some- [14] Alparone S., Gambino S., High precision locations of multiplets what from the previous 02:51-03:00 ones showing ground on south-eastern flank of Mt. Etna (Italy): reconstruction of fault deformation that may be linked to an additional source. plan geometry. Phys. Earth Plan. Int. 2003, 135, 281-289. [15] Lo Giudice E., Patane’ G., Rasa’ R., Romano, R., The structural Acknowledgement: I’m particularly indebted to the tech- framework of Mt. Etna. Mem. Soc. Geol. It., 1982, 23 125-158. [16] Monaco C., Tapponnier P., Tortorici L., Gillot P.Y., Late Quater- nical staff of Tilt Group of INGV-OE for their professional- nary slip rates on the Acireale-Piedimonte normal faults and tec- ism and for ensuring the regular working of the network. tonic origin of Mt. Etna (Sicily) Earth. Planet. Sc. Lett., 1997, 147, I thank Nico Fournier and Karoly Nemeth for their critical 125-139. reading of the manuscript and constructive comments. I [17] Solaro G., Acocella V., Pepe S., Ruch J., Neri M., San- also thank S. Conway for improving the English of this pa- sosti E., Anatomy of an unstable volcano from InSAR: Mul- tiple processes affecting flank instability at Mt. Etna, 1994- per. 2008, Journal of Geophysical Research, 2010, 115, B10405, doi:10.1029/2009JB000820. [18] Alparone S., Cocina O., Gambino S., Mostaccio A., Spamp- inato S., Tuvè T., Ursino A., Seismological features of the Perni- References cana–Provenzana Fault System and implications for the dynamics of northeastern flank of the volcano, J. Volcanol. Geotherm. [1] Wyatt F., Measurements of coseismic deformation in Southern Res., 2013, 251, 16-26. doi:10.1016/j.jvolgeores.2012.03.010. California: 1972-1982, J. Geophys. Res., 1988, 93, 7923-7942. Tilt offset associated with local seismicity Ë 521

[19] Obrizzo F., Pingue F., Troise C., De Natale G., Coseismic dis- [33] Ferro A., Gambino S., Panepinto S., Falzone G., Laudani G., placements and creeping along Pernicana fault (Mt. Etna) in the Ducarme B., High precision tilt observation at Mt. Etna Volcano, last seventeen years: a detailed study of a structure on a vol- Italy. Acta Geophysica, 2011, 59, 3, 618-632. cano. Journal of Volcanology and Geothermal Research, 2001, [34] Gruppo Analisi Dati Sismici., Catalogo dei terremoti della Si- 9, 109–131. cilia Orientale-Calabria Meridionale (1999-2015), INGV Catania, [20] Bonforte A., Carnazzo A., Gambino S., Guglielmino F., Obrizzo 2015, http://www.ct.ingv.it/ufs/analisti/catalogolist.php. F., Puglisi G., The Trecastagni Fault (Mt. Etna, Italy): a multidis- [35] Richter C.F., Elementary Seismology. 1958, Freeman, San Fran- ciplinary study by continuous, discrete and satellite ground de- cisco. formation measurements. Journal of Volcanology and Geother- [36] Frèmont M.J., Malone S.D., High precision relative locations of mal Research, 2013, 251, 42–50. earthquake at Mount St. Helens, Washington. J. Geophy. Res., [21] Gambino S., Bonforte A., Carnazzo A., Falzone G., Ferrari F., 1987, 92, 10223-10236. Ferro A., Guglielmino F., Laudani G., Maiolino V., Puglisi G., Dis- [37] Saraò, A., Panza G.F., Privitera E., Cocina O., Non double couple placement across the Trecastagni Fault (Mt. Etna) and induced mechanisms in the seismicity preceding 1991–1993 Etna vol- seismicity: the October 2009–January 2010 episode. Ann. Geo- cano eruption, Geophys. J. Int., 2001, 145, 319–335. phys., 2011, 54, 414–423. [38] Saraò A. Cocina O., Privitera E., Panza G.F., The dynamic of the [22] Alparone S., Barberi G., Bonforte A., Maiolino V., Ursino A., 2001 Etna eruption as seen with full moment tensor analysis. Evidence of multiple strain fields beneath the eastern flank Geophys. J. Int., 2010, 181, 951-965, doi: 10.1111/j.1365- 246X. of Mt. Etna volcano (Sicily, Italy) deduced from seismic and 2010.04547.x. geodetic data during 2003–2004 Bulletin of Volcanology, 2011, [39] Alparone S., D’Amico S., Gambino S., Maiolino V., Buried ac- http://dx.doi.org/10.1007/s00445-011-0456-1. tive faults in the Zafferana Etnea territory (south-eastern flank [23] Patanè D., Privitera E., Gresta S., Alparone S., Akinci A., Barberi of Mt. Etna): geometry and kinematics by earthquake relo- G., Chiaraluce L., Cocina O., D’Amico S., De Gori P., Di Grazia G., cation and focal mechanisms. Ann. Geophys. 2013, 56 (1), Falsaperla S., Ferrari F., Gambino S., Giampiccolo E., Langer H., http://dx.doi.org/10.4401/ag-5758. Maiolino V., Moretti M., Mostaccio A., Musumeci C., Piccinini [40] Giampiccolo E., D’Amico S., Patanè D., Gresta S., Attenuation D., Reitano D., Scarfì L., Spampinato S., Ursino A., Zuccarello and source parameters of shallow microearthquakes at Mt. L., Seismological features and kinematic constrains for the July- Etna Volcano, Italy, Bull. Seismol. Soc. Am., 2007, 97, 184–197, August 2001 lateral eruption at Mt. Etna volcano, Italy. Ann. of doi:10.1785/0120050252. Geophysics, 2003, 46, 599-608. [41] Tuvè T., D’Amico S., Giampiccolo E., A new MD-ML relationship [24] Cauchie L., Saccorotti G., Bean C. J., Amplitude and recurrence for Mt. Etna earthquakes (Italy). Annals of Geophysics, 2015, 58, time analysis of LP activity at Mount Etna, Italy, J. Geophys. Res. 6, doi:10.4401/ag-6830S0657. Solid Earth, 2015, 120, 6474–6486, doi:10.1002/2015JB011897. [42] Bozzano F., Gaeta M., Lenti L., Martino S., Paciello A., Palladino [25] Patanè D., Cocina O., Falsaperla S., Privitera E., Spampinato S., D.M., Sottili G., Modeling the effects of eruptive and seismic ac- Mt. Etna volcano: a seismological framework. In: S. Calvari, A. tivities on flank instability at Mount Etna, Italy. Journal of Geo- Bonaccorso, M. Coltelli, C. Del Negro and S. Falsaperla (Eds.), physical Research: Solid Earth, 2013, 118(10), 5252-5273. The Mt. Etna Volcano. AGU, Washington, D.C., 2004, 147–165. [43] Aki K., Generation and propagation of G waves from Niigata [26] Gambino S., Mostaccio A., Patanè D., Scarfì L., Ursino A., earthquake of June 16, 1964, II, Estimation of earthquake mo- High-precision locations of the microseismicity preceding the ment, released energy, and stress-strain drop from the G wave 2002–2003 Mt. Etna eruption. Geophys. Res. Lett. 2004, 31 (18), spectrum, Bull. Earthquake Res. Inst. Univ. Tokyo, 1966, 44, 73- L18604 10.1029/2004GL020499. 88. [27] Bonaccorso A., Bonforte A., Gambino S., Twenty-five years [44] Okada Y., Surface deformation due to shear and tensile fault in of continuous borehole tilt and vertical displacement data at half-space. Bull Seismol Soc Am 1985, 75:1135-1154. Mount Etna: Insights on long-term volcanic dynamics, Geophys. [45] King C.Y., Nason R.D., Burford R.O., Coseismic steps recorded Res. Lett., 2015, 42, doi:10.1002/2015GL066517. on creep meters along the San Andreas fault, J. Geophys. Res., [28] Behncke B., Neri M., The July–August 2001 eruption of Mt. Etna 1977, 82, 1655-1662. (Sicily). Bull. Volcanol. 2003, 65, 461–476. [46] Allen R.V., Tiltmeter observations near a large earthquake, Bull. [29] Bonforte, A., Gambino S., Neri M., Intrusion of eccentric dikes: Seism. Soc. Am. 1978, 68, 855-857. the case of the 2001 eruption and its role in the dynamics of Mt. [47] Bonaccorso A., Falzone G., Gambino S., An investigation into Etna volcano. Tectonophysics, 2009, 471, 78–86. shallow borehole tiltmeters, Geophys. Res. Lett. 1999, 26, 11, [30] Gambino S., Continuous dynamic response along a pre-existing 1637-1640. structural discontinuity induced by the 2001 eruption at Mt [48] Hill D.P.,Langbein J.O., Prejean S., Relations between seismicity Etna. Earth Planets Space, 2004, 56, 447-456. and deformation during unrest in Long Valley Caldera, Califor- [31] Westphal J.A., Carr M.A., Miller W.F., Dzurisin D., Expendable nia, from 1995 through 1999, J. Volcanol. Geotherm. Res., 2003, bubble tiltmeter for geophysical monitoring, Review of Scien- 127, 175–193,doi:10.1016/S0377-0273(03)00169-0. tific Instruments, 1983, 415-418. [49] Llenos A.L., McGuire J.J., Ogata Y., Modeling seismic swarms [32] Dzurisin D., Electronic tiltmeters for volcano monitoring: triggered by aseismic transients, Earth Planet. Sci. Lett., 2009, lessons from Mount St. Helens, in Monitoring volcanoes: tech- 281, 59–69. niques and strategies by the staff of the Cascades observatory, [50] Cannavò F., Gambino S., Puglisi B., Velardita R., Modeling 1980-90, Ewert, J.W., and D.A. Swanson, Eds, U.S.G.S. Book, ground deformation associated with the destructive earth- 1992, 69-83. quakes occurring on Mt. Etna’s southeastern flank in 1984. Nat. Hazards Earth Syst. Sci. Discuss., 2016, doi:10.5194/nhess- 522 Ë Salvatore Gambino

2015-312. [53] Siniscalchi A., Tripaldi S., Neri M., Balasco M., Romano G., [51] Murru M., Montuori C., Wyss M., Privitera E., The location Ruch J., Schiavone D., Flank instability structure of Mt. Etna in- of magma chambers at Mt. Etna, Italy, mapped by b-values. ferred by a magnetotelluric survey, J. Geophys. Res., 2012, 117, Geoph. Res. Lett., 1999, 26, 2553-2556. B03216, doi:10.1029/2011JB008657. [52] Bonforte A., Bonaccorso A., Guglielmino F., Palano M., Puglisi G., Feeding system and magma storage beneath Mt. Etna as revealed by recent inflation/deflation cycles, J. Geophys. Res., 2008, 113, B05406, doi:10.1029/2007JB005334.