Instruction manuals

A multidisciplinary study of an active fault crossing urban areas: The Trecastagni Fault at Mt. Etna (Italy)

Description
A multidisciplinary study of an active fault crossing urban areas: The Trecastagni Fault at Mt. Etna (Italy)
Published
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A multidisciplinary study of an active fault crossing urban areas: The TrecastagniFault at Mt. Etna (Italy) A. Bonforte  a, ⁎ , A. Carnazzo  b , S. Gambino  a , F. Guglielmino  a , F. Obrizzo  c , G. Puglisi  a a Istituto Nazionale di Geo  fi sica e Vulcanologia, Sezione di Catania - Osservatorio Etneo, Piazza Roma 2, 95123 Catania, Italy b Provincia Regionale di Catania,Via Novaluce,67/A - 95030 Tremestieri Etneo (CT), Italy c Istituto Nazionale di Geo  fi sica e Vulcanologia,Osservatorio Vesuviano, Via. Diocleziano 328, 80124 Napoli, Italy a b s t r a c ta r t i c l e i n f o  Article history: Received 11 November 2011Accepted 5 May 2012Available online 24 May 2012 Keywords: InSAR LevellingFaultVolcano-tectonicsEarthquakes The Trecastagni Fault is a NNW – SSE tectonic structure in the densely inhabited southern  fl ank of Mt. Etna,characterisedby evident morphological scarps and movements of normal and right-lateral type thatdirectly af-fect roads and buildings. The fault is affected by continuous dynamics with intermittent accelerationsaccompa-nied with shallow seismicity. It has an important role in the instability affecting Mt. Etna's south-eastern  fl ankand represents part of the southern boundary of the unstable sector. The motion of the fault between 2005and2011hasbeenanalysedbyusingamulti-disciplinaryapproachinvolvingterrestrialandsatellitegroundde-formation data. Active monitoring systems able to investigate the fault in detail are extensometers, a levellingnetwork and InSAR. Two episodes of acceleration were recorded at the end of 2009 and during 2010. Data evi-dences that the acceleration episodes affected only portions of the fault and that stress may accumulate andbeperiodicallyreleased.Althoughbothmagmaticprocesses(in fl ationorintrusiveepisodes)and fl ankdynamicsin fl uencetheoccurrenceoftheTFaccelerationepisodes,thedraggingeffect oftheoverallseawardslidingofthesouth-eastern  fl ank is evident and it causes the subsidence of the hangingwall, accumulating stress on the faultthat is periodically seismically released.© 2012 Elsevier B.V. All rights reserved. 1. Introduction Most volcanoes are affected by  fl ank instability, generally due tothe combination of gravitational force, magmatic intrusion and inter-action with the substratum morphology and lithology. Flank instabil-ity could lead to a slow and not necessarily symmetrical spreading of the volcano, as well as to sudden and dramatic collapse of the edi fi ce.In any case,  fl ank motion produces complex faulting on the volcano,fromthesummittowardsitsperiphery;upper partsof thesefaultsys-tems canrepresent often preferentialways for lateral dyke intrusions,forming the main rift zones of the volcano, while lower parts workmainly as volcano-tectonic faults, bordering the collapsing sectors.Studying and understanding the faults controlling the sector collapsecan provide fundamental information in modelling and monitoringthe volcano dynamics, providing useful data improving the knowl-edge of the interaction and feedback between magmatic and tectoniccontrol of the volcanic and seismic activity affecting the  fl anks of avolcano.In this paper we investigate the 2005 – 2011 displacements andseismicityof theTrecastagni faultthat runsacrossa denselyinhabitedarea on the SE  fl ank of Mt. Etna. This fault represents an importantfeature, related to the southern boundary of the sliding  fl ank of thevolcano; ground deformation data (Solaro et al., 2010; Bonforte etal., 2011a, 2011b) evidence how it is one of the most active structuresdecoupling the mobile sector on the southern  fl ank. Furthermore, itcrosses several villages on this side of the volcano, producing signi fi -cant damage due to its shallow seismic activity and to its continuousaseismic creep. We analyse its behaviour in the framework of the fl ank dynamics and volcanic activity of Mt. Etna through a multi-disciplinary approach.Flank dynamics at Mt. Etna seem to play an important role as a “ trigger ”  of the volcanic activity (e.g. Walter et al., 2005; Allard etal., 2006; Bonforte et al., 2007a, 2007b, 2008). Mt. Etna volcano is lo-cated at the intersection of two main regional fault systems, havingNNW – SSE and NE – SW trends respectively (Fig. 1) producing a gener-al downthrow of the easternmost side of the volcano. Furthermore,while the northern part of the edi fi ce lies over the rocky tectonic unitsof the Maghrebian orogenic chain, the southern  fl ank lies on clay units fi lling the foredeep. All these features contribute to create a complexstructural framework on the edi fi ce, induced by the composite inter-action between regional stress, gravity forces and dike-inducedrifting. All these factors contribute to produce the slow sliding of eastern and south-eastern  fl ank of the volcano ( fl ank dynamics),whose main evidence is the many fault systems cropping out inthis area, the frequent seismic activity and the intense ground defor-mation (Bonforte and Puglisi, 2006).  Journal of Volcanology and Geothermal Research 251 (2013) 41 – 49 ⁎  Correspondingauthorat:IstitutoNazionalediGeo fi sicaeVulcanologia – OsservatorioEtneo, Piazza Roma 2, 95123 Catania, Italy. Tel.: +39 95 7165800; fax: +39 95 7165826. E-mail address:  alessandro.bonforte@ct.ingv.it (A. Bonforte).0377-0273/$  –  see front matter © 2012 Elsevier B.V. All rights reserved.doi:10.1016/j.jvolgeores.2012.05.001 Contents lists available at SciVerse ScienceDirect  Journal of Volcanology and Geothermal Research  journal homepage: www.elsevier.com/locate/jvolgeores  On the north-eastern side, where the edi fi ce is buttressed by arocky high substratum, the boundary of the sliding sector of the vol-cano is represented by the highly active Pernicana – Provenzana FaultSystem (PFS). Evidence of the  fl ank dynamics abruptly disappearsnorthward of this important feature (Bonforte et al., 2007c;Guglielmino et al., 2011). On the south-eastern side, where the thin-ner volcanic pile lies on a lower and clay substratum, the boundary isless de fi ned because all evidence gradually decay (Borgia et al., 1992;Lo Giudice and Rasà, 1992; Froger et al., 2001): the deformationsgradually decrease south-westward across the entire southern sideof the volcano; the seismicity becomes shallower and/or less intense;the fault systems are less continuous and/or evident. A signi fi cantpart of the ground deformation is distributed on different fault sys-tems that progressively accommodate the motion of the eastern fl ank as initially evidenced also by early long-period SAR interfero-grams (Borgia et al., 2000; Froger et al., 2001). Among the different known faults involved in the south-eastern boundary we may men-tion, from West to East, the Ragalna Fault system (Neri et al., 2007), the Mascalucia – Tremestieri Fault and the Trecastagni – TremestieriFault (TF). These faults are characterised by evident morphologicalescarpments, sometimes masked by the presence of cones and lavas.In order to investigate the ground deformation pattern associatedwith the Trecastagni volcano-tectonic structure, a multi-disciplinaryapproach is presented here. It represents a fundamental improve-ment in the investigation of an active fault, since it comprises and in-tegrates in-situ and remote sensing data, in order to overcome thedif  fi culties deriving from the intensely urbanised areas crossed bythis fault. Remote sensing data is the most suitable for this kind of en-vironment; in addition, as a ground truth, we integrated satellite databydesigningandinstallingalevellingnetworkthatprovides themostprecise vertical data and is not conditioned by poor sky visibility likeGPS is;  fi nally, we added the data coming from extensometers inordertoachieve themaximum detail in the timeseriesof thefault mo-tion.Bythis data combination,it is possibleto investigate animportantactive fault, even where it crosses urban areas that hide its  fi eld evi-dence and prevent some in-situ measurements and installations. 2. The Trecastagni Fault The Trecastagni Fault (TF) is a discontinuity that develops in thesouthern fl ank of Mt. Etna, between the Trecastagni and San Giovannila Punta villages. This is an active structure with an approximatelyNNW – SSE trend (Fig. 1) characterised by morphological escarpmentsand very shallow seismicity. The effects of the activity of the TF(creep) are visible on much of the road 8/III.The seismicity of the TF is characterised by very shallow earth-quakes with typical focal depths of 1 – 2 km. Evident co-seismic sur-face faulting occurred along the fault scarp in September 1980 andin November 1988 (Azzaro, 1999). Hollows appeared in agriculturalland and fractures offsetting buildings, boundary walls and the road8/III, were observed after the local shocks on September 16 and 20,1980 which displaced the northernmost and the central segments of the fault respectively and also after the 21 November 1988 earth-quake (Azzaro, 1999). Three other similar episodes on 26 May 1903, 20 July 1917 and 17 February 1955, have been reconstructed by his-torical research (Azzaro and D'Amico, 2008). In recent years, someminor shallow earthquakes have been felt by the local residents;in particular on 15 October 2009, at 00:52 GMT, M=2.1 and on 29October 2010, at 01:37, M=2.2 (http://www.ct.ingv.it/ufs/analisti/catalogolist.php).Data on deformations across TF are relatively sporadic, until 2005.Froger et al. (2001) estimated a deformation rate of about 4 – 6mm/yearby analysing synthetic aperture radar (SAR) interferograms, coveringthe 1996 –  January 1998 Mt. Etna in fl ation period. Recent time series Fig. 1.  Surface faults sketch map of Mt. Etna. Inset map shows the main regional fault systems: MF = Messina – Fiumefreddo line, ME = Malta Escarpment, RDN = Ripe della Naca.Grey lines de fi ne the sliding sector.42  A. Bonforte et al. / Journal of Volcanology and Geothermal Research 251 (2013) 41 – 49  analysisofSARdataspanningfrom1995to2000byusingthePermanentScatters technique (Bonforte et al., 2011a) and from 2003 to 2008 by using the Small Baseline technique (Solaro et al., 2010), quanti fi ed anaverage downthrow of the eastern side of the TF at a rate of about4 – 5 mm/year and a negligible East – West movement.The intrusion of July 2001 (2001 eruption) was very important inthe recent dynamics of Mt. Etna (e.g. Bonforte et al., 2009) since it was produced by a rapid eccentric intrusion that displaced the entireedi fi ce, causing the  fi rst lateral eruption after a 9-year in fl ation andinducing an acceleration of the deformations and seismic activityalong the Ionian coast in the following months (Puglisi et al., 2008).During the 2002 – 03 Mt. Etna eruption, Neri et al. (2004), by mea- suring dislocations on the ground fractures, claim to have measureddisplacements of more than 2 cm after a Md=2.6 (duration magni-tude) shallow earthquake recorded on 26 November, 2002 inter-preted as the migration of the ground deformation induced by theeruptive intrusion along the NE Rift that  fi rst affected the PernicanaFault. 3. Ground deformation techniques and data The INGV-CT continuous geodetic networks (Tilt and GPS, Fig. 2a)operating on Mt. Etna, despite the overall high density of points de-voted to ef  fi ciently monitoring the dynamics of the volcano (Palanoet al., 2010), do not allow a local and detailed study of the Trecastagnifault. Benchmarks of the GPS discrete network, enable a detailedstudy and modelling of the structural framework of the volcano (seeBonforte et al., 2008 and references therein), but are not close enoughto this sector of the fault to allow an investigation needed for a de-tailed study on a local scale. For this reason, INGV installed various in-strumentation to study and monitor the displacement and grounddeformation associated to this hazardous portion of the Trecastagnifault. At present, the systems that are able to investigate the fault of Trecastagni in detail are the extensometers installed in 2005, thelevelling network installed in 2009 (Fig. 2b) and InSAR remote sens-ing techniques.  3.1. Remote sensing measurements 3.1.1. Permanent scatterers (ERS and ENVISAT time series) Permanent scatterer technique (PS-InSAR) is a particular algo-rithm developed for processing repeated data acquired by SyntheticApertureRadar(SAR)on boardspacecrafts; it represents asubstantialimprovement to classical differential interferometry approaches thatare affected byseveraladditional factorswhichadd noise to the inter-ferometric phase, such as temporal and geometrical decorrelationproblems, topographic effects, orbit errors, as well as atmospheric ar-tefacts (Zebker and Villasenor, 1992; Massonnet and Feigl, 1995;Zebker et al., 1997; Bonforte et al., 2001; Onn and Zebker, 2006).Atmosphericartefactscanbeestimatedandthenremovedbycombiningdata from long time series of SAR images (Ferretti et al., 2000, 2001;Colesanti et al., 2003).FromPS-InSARmapsanalysedbyBonforteetal.(2011a),byconsid-ering all available ERS-1 and ERS-2 images acquired on both ascendinganddescendingorbitsacquiredfrom1995to2000,aNNW – SSEdiscon-tinuity in the ground velocity  fi eld (named PSF6 in Bonforte et al.,2011a)wasdetectedonthesouth-easternsideofthevolcano,followingthe path of the Trecastagni fault, but extending from lower altitude(about 300 m) up to medium-high  fl ank (about 1500 m), towards thearea affected by large ground fracturing (gaping cracks up to morethan 1 m; Ferrucci et al., 1993) during the 1989 eruption and where a similar dislocation with the same azimuth was already modelled byGPS and InSAR data during the 2001 one (Bonforte et al., 2004, 2009;Puglisietal.,2008).Thisfaultshowsamainverticalkinematics,produc-ingthe strongesteffect on the vertical velocity pattern, withanevidentdownthrow of the hangingwall (on the eastern side) at a rate of about4 mm/year (±about 1 mm/year, from Bonforte et al., 2011a) with re- spect to the footwall. Subsidence continues to increase eastwards awayfrom the structure, reaching a maximum rate of almost 1 cm/year. Thefault produces a minor increase in the eastwards velocity on its east-ern side (hangingwall) evidencing also a minor extension of thestructure. Solaro et al. (2010), by using a similar approach to recon-struct the time series on coherent pixels from ERS and ENVISATdata, de fi ne the same kinematics of the TF fault. Furthermore, bycomparing two different time periods, 1994 to 2000 and 2003 to2008, they detect different rates of deformation before and afterthe 2001 and 2002 – 2003 lateral eruptions; in particular, theyfound a higher subsidence rate of their sector 5 (East of the TF)after 2003.To extend the analysis to a more recent time period in order tohave a time-consistent geodetic dataset, we also investigated a newPS time series by processing all the available ENVISAT ascending im-agery from January 2007 to January 2010 with the StaMPS package(Hooper, 2008). The ground velocity for this period is shown in Fig. 3 and has been measured only along the Line Of Sight (LOS) of the sensor that looks from West to East with an incidence angle of about 23° from the zenith. The entire map of the ground velocity(Fig. 3a) shows a very slight average de fl ation of the upper part of the volcano with an important lowering of its uppermost NE  fl ank.More in detail, on the lower SE  fl ank here investigated (Fig. 3b), thisdataset con fi rms the kinematics observed by the previous one, witha sharp increase of the motion away from the sensor on the easternside of the TF; this velocity pattern is in agreement with a loweringand/or eastward motion of the hangingwall of the fault. The deforma-tion rate is rather higher than the 1995 – 2000 period (Bonforte et al.,2011a) and this could be due to the fact that the LOS velocities are Fig. 2.  Maps of the ground deformation monitoring networks on Mt. Etna (a), indicating also the location of the two earthquakes affecting the lower southeastern fl ank, and a detailon the Trecastagni fault (TF) area (b) with the new levelling network installed in 2009. PSF (Permanent Scatterers Features), as de fi ned by Bonforte et al. (2011a) are reported asbrown lines.43  A. Bonforte et al. / Journal of Volcanology and Geothermal Research 251 (2013) 41 – 49  sensitive to both horizontal (E – W) and vertical components of groundmotion and/or to an accelerated dynamics of the fault in 2007 – 2010.  3.1.2. DInSAR (Radarsat pair) In order to image the short-term ground deformation occurring intherecentperiod,weproducedaRadarsat2C-bandimage.Weprocessedan interferogram (Fig. 4) by using the 29 June 2010 – 31 January 2011pair, acquired in image mode, beam F3 (incidence angle=41.8 – 44.3°),on descending orbit geometry (looking from East to West). The proce-dure used for the generation of the differential interferogram is the so-called  “ two-pass interferometry ”  (Massonnet and Feigl, 1998), and we used the SarScape package developed by SARMAP to process it.During this 6-month period, the interferogram (Fig. 4a) shows ade fl ation of the upper part of the volcano (above about 1500 ma.s.l.). In addition to this wide deformation pattern, a local deforma-tion is detected on the southern part of the TF. A LOS ground motionaway from the sensor by almost one fringe (one fringe correspondsto28 mm for Radarsat2 C-Band, while the overall uncertainty in thephase estimation is in the order of few millimetres) affects a 4 km 2 area of the north-western part of San Giovanni la Punta village, onthe hangingwall close to the fault, where the southern side of thelevelling network measured a jump of about 25 mm crossing thefault during the 2010 – 2011 period (Fig. 4b). The LOS deformationdetected by the descending interferogram can be due to subsidenceand/or westwards motion of the ground; considering the known ki-nematics of the TF, Radarsat data con fi rm that an important subsi-dence affected that area with the usual normal behaviour of the fault.  3.2. Ground based geodetic measurements 3.2.1. Extensometers From 2005, the INGV and the Provincia di Catania, began to imple-ment a local monitoring system, aimed chie fl y at improving the directmeasurements along faults and fractures (Gambino, 2004; Neri et al.,2004), quantifying the creeping phenomena across the local roads andthus to relate this deformation with the current volcanic activity of Mt.Etna.Atpresent,themonitoringoftheTFconsistsoftwocontinuouswire extensometers and a system for periodic direct measurements Fig.3. Ground velocities calculated onpermanent scatterersbytheSTaMPS package,using the ascendingimagery ofENVISAT spacecraft from2007 to2010. PSFsfromBonforte et al.(2011a) are reported with brown lines. Fig. 4. Radarsat2 descending interferogram, from June 2009 to January 2011 showing a wide de fl ation of the volcano (a) and a very local subsidence close to the southern part of theTrecastagni fault (b). In the zoom at the right (b) on a Google map, the levelling network is also reported by red circles; the dashed circle encompasses the subsiding area. PSFs fromBonforte et al. (2011a) are reported with brown lines.44  A. Bonforte et al. / Journal of Volcanology and Geothermal Research 251 (2013) 41 – 49  acrossthediscontinuitiesinthecentralandnorth-centralsectorsof thefault (Fig. 2).We installed (Fig. 5a, b) two continuous wire extensometers madeby Sisgeo (Mod. D241A20) at two stations (ET1 and ET2). Each stationis equipped with a compact datalogger programmed for 48 data/daysamplingandincludesacquisitionofthedisplacementandgroundtem-perature (more details in Carnazzo et al., 2006; Gambino et al., 2011).The ET1 was installed in May 2005 directly on a fractured concretestructure (Fig. 5a). ET2 was set up in 2007 on the ground (Fig. 5b) after creating an 8 m long and 30cm deep trench and sliding the wireinside a PVC pipe. The two stations measure the relative displacementsperpendicular to the fracture (the extensional component).In Fig. 6, we report May 2005 – Feb 2011 data recorded at the twoextensometers. Displacement data are characterised by a low dailynoise ( b 0.1 mm). Data recorded between May 2005 and September2009 highlight an opening trend of about 2 – 3 mm/year, while fromOctober 15, 2009 data show a sharp increase in the trend that cumu-lated more than 2 mm in the  fi rst 15 days and a total of 4 mm by theend of January 2010. From February 2010 to February 2011, the trendseemstoreturnto thesamevaluemeasured beforeOctober2009.TheET2 instrument, which has been affected by several technical and en-vironmental problems, appears to record a variation of 5 – 6 mm dur-ing the period October 2009 –  January 2010.  3.2.2. Levelling  In order to have a precise ground validation of the vertical compo-nent of velocities measured by Solaro et al. (2010) and Bonforte et al. (2011a), a new levelling network was set up in November 2009 overthe area showing the maximum vertical slip (Obrizzo et al., 2001,2004; Bonforte et al., 2011b). The network (Figs. 2b, 7) has a loop (about 6 km long), crossing the fault two times on the northern andsouthern sides, and three open lines. Three branches start off fromthe main loop, one extending westwards (from T3 to T0) towardsthe stable part and two extending eastwards (from T10 to T19) andsouth-eastwards (from T9 to T17) towards the area showing themaximum subsidence rate. The entire levelling network is 9 kmlong, distributed along the roads close to the geological structure,and consists of 20 benchmarks. Surveys were carried out with anelectronic Leica DNA03 level, an optical level Wild NA2 equippedwith parallel-plate micrometre (resolution of 0.01 mm) and invarrods; instruments were calibrated every day before the measurements.Maximumaccepteddiscrepancy(inmm)betweentheheightdifferencesmeasuredduringforwardandbackwardpathsis±2.5(L) 1/2 ,L(inkm)isthe length of levelling section; the maximum error (in mm) allowed fortheclosureoftheloopis±2.0(C) 1/2 ,C(inkm)isthelengthofthecircuit(Tables 1, 2). (Commissione Geodetica Italiana, 1975; Bomford, 1971; Bonforte et al., 2011b).The measurements were carried out in November 2009, March2010 and March 2011. The reference benchmark used to computethe height variations is the benchmark T1, which lies on the westernside of the fault, at a distance of about 0.5 km from it; the position of this benchmark was chosen from PS maps, in a stable area that wasunequivocally outside the subsiding block bordered by the fault.Fault trace intersects the levelling route at benchmarks T5 – T6 onthe southern side and at benchmarks T14 – T15 – T16 on the northernone (Bonforte et al., 2011a). By comparing the results of the November 2009 survey with thoseof the March 2010 one (Fig. 7a, Table 1), the slip related to the Trecastagni fault can be detected. In this 4-month comparison, aclear jump of about 1 to 2 mm (±1 mm) affects stations across thefault. Indeed, while all benchmarks lying on the western side of thefault show no variation (less than 0.2 mm±1 mm) in their verticalpositions from November 2009 to March 2010, those lying just eastof the structure suddenly show a slight subsidence of about 1 mm(±1 mm) on the southern part of the network up to 2 mm on thenorthern part. As revealed by PS-InSAR velocity maps (Bonforte etal., 2011a), vertical motion gradually increases towards east, awayfrom the fault trace. From levelling data referring to this 4-month in-terval and reported in Table 1, maximum subsidence of 4.5 mm(±1.5 mm) has been measured on the north-eastern part of the net-work, resulting in a subsidence rate of about 13.5 mm/year.The subsequent comparison covers a longer time span, betweenthe March 2010 and March 2011 surveys, and provides a more inter-esting pattern of ground deformation (Table 1, Fig. 7b). Due to the Fig. 5.  Photos illustrating the two extensometers (A: ET1, B: ET2) installed across theTrecastagni fault.Redrawn from Gambino et al. (2011). Fig.6. Plotof theextension measured atthe twoextensometers from 2005to 2011.TheET1 data has been  fi ltered using a linear distance – temperature regression (more de-tails in Gambino et al., 2011).45  A. Bonforte et al. / Journal of Volcanology and Geothermal Research 251 (2013) 41 – 49
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks