The influence of erosional processes on the visibility of Permanent Scatterers Features from SAR remote sensing on Mount Etna (E Sicily)

Geomorphology 198 (2013) 128–137

Contents lists available at ScienceDirect

Geomorphology

j ourna l homepage: www.e lsev ie r .com/ locate /geomorph

The influence of erosional processes on the visibility of Permanent Scatterers Featuresfrom SAR remote sensing on Mount Etna (E Sicily)

S. Catalano a,⁎, A. Bonforte b, F. Guglielmino b, G. Romagnoli a, C. Tarsia a, G. Tortorici a

a Dipartimento di Scienze Biologiche, Geologiche ed Ambientali – Sezione di Scienze della Terra, Università di Catania, C.so Italia, 55, 95129 Catania, Italyb Istituto nazionale di Geofisica e Vulcanologia – Osservatorio Etneo, Piazza Roma, 2, 95125 Catania, Italy

⁎ Corresponding author. Tel.: +39 095 7195714; fax:E-mail addresses: [email protected] (S. Catalano), bo

[email protected] (F. Guglielmino), g.romagnoli@[email protected] (C. Tarsia), [email protected] (G. T

0169-555X/$ – see front matter © 2013 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.geomorph.2013.05.020

a b s t r a c t

Article history:Received 1 October 2012Received in revised form 21 May 2013Accepted 22 May 2013Available online 29 May 2013

Keywords:DInSARGround deformationActive tectonicsMount EtnaEastern Sicily

Analysis of 1549 DInSAR interferograms, covering the period from 2003 to 2010, has highlighted significant mo-tion along the entire set of the active faults identified by advanced DInSAR analyses (i.e. Permanent ScatterersFeatures, PSF), affecting the Mount Etna volcano, in eastern Sicily. In the analysed period, the absence of signif-icant seismicity producing co-seismic ground deformation suggests that the overall deformation that has beenrecognized on the interferograms is to be associated with interseismic, almost continuous creep which is, welldocumented along most of the active faults. According to field evidence, the structures should accumulate dis-placements resulting in their permanent visibility on the interferograms, progressively increases through time.This expected behaviour has been recognised only for part of the entire set of structures. Other tectonic features,in fact, show episodic appearances, alternating with periods of absence of ground displacement on the interfero-grams, simulating a stick-slip mechanism of deformation, conflicting with field evidence. This apparently incon-gruous behaviour can be interpreted as the result of topographic changes due to the combination of the tectonicdisplacements with related amounts of the differential erosion and deposition across the fault line. The compar-ison between the history of the appearances and the monthly rainfall in the region seems to demonstrate thatthese structures appear when one of the two interacting processes governing the topographic changes aroundthe fault, i.e. tectonic vs. erosional, prevails over the other. Otherwise, the same structures are not evident onthe interferograms when the two components are in balance.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Mount Etna, in eastern Sicily, is one of the most studied volcanoes byInterferometry SAR (InSAR) spatial techniques. This method, achievablewith active sensors emitting microwaves, can detect displacements anddeformations with centimetric to millimetric precision, depending onthe applied specific InSAR technique (Massonnet and Feigl, 1998). Inthis paper we discuss the results of the ground deformation measure-ments by SAR remote sensing at Mount Etna, from the 2003 to the2010. The analysed dataset consists of Differential Interferometry SAR(DInSAR) interferograms generated by ENVISAT images. The study hasbeen focused on the active tectonic lineaments of Mount Etna (Monacoet al., 1997, 2005; Azzaro, 1999; Catalano et al., 2004; Azzaro et al.,2012), which have been defined and characterized as “PermanentScatterers Features” (PSFs) by Bonforte et al. (2011), on the basis ofERS-1/2 SAR data acquired from 1995 to 2000, and recently summa-rizedwith all known geometric and kinematic parameters in a GIS data-base by Barreca et al. (2013). The study aims (i) to verify the activity of

+39 095 [email protected] (A. Bonforte),nict.it (G. Romagnoli),ortorici).

l rights reserved.

the PFSs in the period of 2003–2010; (ii) to eventually detect new inter-ferometric features; and (iii) to analyse the short- and long-termbehav-iour of these discontinuities in the ground deformation field, in order totest the correlation of the history of their visibility on the SAR imageswith the dynamic processes active in the region during the consideredtime interval.

2. Structural setting of Mount Etna

The Mount Etna volcano, in eastern Sicily, lies on the northern edgeof the Nubia Plate, where the E-W trending Nubia-Eurasia convergingmargin (Serpelloni et al., 2007) is interrupted by an incipient N-S ori-ented rift zone, the Siculo-Calabrian Rift Zone (SCRZ) (Catalano et al.,2008, 2010, 2011) (Fig. 1). The earlier fissural eruptions of tholeiiticproducts occurred along a main NW-SE oriented belt located alongthe southern border of the modern stratovolcano (Gillot et al., 1994;Branca et al., 2007), which was controlled by a main dextral shear zone,dissecting the external areas of the Sicily collisional belt (Catalano et al.,2004). This main shear zone represents a transform belt which separatesthe contractional region of the Hyblean Plateau (Catalano et al., 2010,2011), to the southwest, from the active extensional domains of MountEtna to the northeast (Catalano et al., 2004, 2008), where the activeNa-alkaline volcanism of Mt. Etna is confined at the footwall of the rift

Fig. 1. Geological sketch map of the eastern Sicily. Key: (1) Holocene deposits (a), Late Würmian-Holocene alluvial fan (b); (2) Etnean Lavas: recent alkaline products (80 ka toPresent) (a), ancient alkaline products (180–100 ka) (b) (chronological data from Gillot et al., 1994); (3) sub-alkaline products of Mt. Etna volcanic district (320–200 ka) (chronologicaldata from Gillot et al., 1994); (4) Middle Pleistocene deposits of the Simeto and Scordia-Lentini Graben and Late Quaternary marine terraces; (5) Early- Middle Pleistocene deposits of theGela Foredeep; (6) UndifferentiatedMeso-CenozoicMaghrebian allochthonous units; (7) Undifferentiated sequences of theHyblean Plateau; (8) Pliocene- Early Pleistocene fault; (9)MiddlePleistocene extensional fault (Simeto and Scordia-Lentini Graben); (10) Late Quaternary and active tectonics and volcano tectonics: strike-slip fault (a), normal faults of the “Siculo-CalabrianRift Zone” (b), dry and eruptive fissure (c); (11) main transform fault zone; (12) maximum horizontal extension in the Etnean area and in the Ionian offshore; (13) maximum horizontalcompression in the Catania Plain area. The inset a refers to the relative motion between the plates.

129S. Catalano et al. / Geomorphology 198 (2013) 128–137

130 S. Catalano et al. / Geomorphology 198 (2013) 128–137

zone (Fig. 1). In the active tectonic picture of Mount Etna, the NW-SEoriented Mascalucia-Tremestieri Fault and the more recently definedBelpasso-Ognina Fault (namely PSF 8 in Bonforte et al., 2011, 2012a; seeFig. 2b) together with the adjacent E-W oriented San Gregorio-AcitrezzaFault (Monaco et al., 2010) (Fig. 2c) could represent the response of thesouthern part of the volcanic edifice to the kinematics of this transform

Fig. 2. PSFs and related active tectonics of Mount Etna: a) Ground LOS velocity (mm/y) mapsal., 2011, mod.); b) PSFs over the active fault traces; c) detail of the active faults of the east

zone. The current deformation processes of the volcano, shown by geo-detic and geological surveys, have been mainly referred to tectonics,dike intrusions, gravitational spreading and deeper pressure sources andinteraction of all these phenomena controlling the volcano stability andactivity (Rasà et al., 1996; Monaco et al., 1997; Froger et al., 2001;Houlie et al., 2006; Bonforte et al., 2008; Bonaccorso et al., 2011).

for ascending and descending view, in which the PSFs are highlighted (from Bonforte etern flank of Mount Etna (from Azzaro, 1999; Monaco et al., 2010, mod.).

131S. Catalano et al. / Geomorphology 198 (2013) 128–137

Recently, Bonforte et al. (2011) carried out a study of active tec-tonic features of Mt. Etna applying the Permanent Scatterers (PS)method, an improvement to the classic Differential SAR Interferome-try (DInSAR) approaches (Curlander and McDonough, 1991; Ferrettiet al., 2001). The authors, produced detailed both ascending and de-scending ground velocity maps (Fig. 2a) to discriminate, from theircombination, vertical and horizontal (E-W) velocities. The two resultingmaps are characterised by the occurrence of distinct linear interrup-tions and jumps in the ground deformation pattern that have been de-fined as “Permanent Scatterers Features” (PSFs) (Bonforte et al., 2011).The PSFs, separating blocks with different velocities, are often associat-ed with the main active faults of Mount Etna (Fig. 2), thus providing asignificant tool for defining tectonic deformation at both regional andlocal scale of the volcano.

PSF1 corresponds to the arc-shaped lineament that includes theNE-Rift and the Pernicana Fault, along the north-eastern flank ofMt. Etna, abruptly closing the ESE-wards motion of the mobile flank ofthe volcano. The Pernicana Fault is the most active tectonic feature ofMt. Etna, extending for about 20 km, showing a 20–30 mm/y left-lateraldisplacement-rate (Azzaro et al., 1998; Bonforte et al., 2007). Thewesternhalf of this fault is marked by a well-defined south-facing scarp, dueto a small dip-slip component, that interrupts the continuity of thenorth-eastern slope of the volcano.

Southwards the Timpe Fault system represents the most impressivemorpho-tectonic feature of the Etnean region (Azzaro et al., 2012;Bonforte et al., 2012a). This fault system is composed of distinct fault seg-ments, in the lower eastern sector of the volcano. The system includesthe 18-km-long NNW-trending, east-dipping fault (Fig. 2c), extendingfrom the Timpa of Acireale (Acireale Fault; Monaco et al., 1997; Azzaro,1999) to the Timpa of Moscarello (Moscarello Fault by Azzaro, 1999;S. Alfio Fault by Monaco et al., 1997). Further to the east, the 7 km-long,east-dipping S. Leonardello Fault (Monaco et al., 1997; Azzaro, 1999;Azzaro et al., 2012) runs parallel to the previous one and, together withthe northern termination of the Acireale Fault, corresponds to the PSF2by Bonforte et al. (2011). These east-dipping normal faults are faced bythe antithetic Trepunti Fault segments, located immediately to theeast of the S. Leonardello Fault (Fig. 2). To the south, the Timpe FaultSystem overlaps, with a right-stepping arrangement, the Nizzeti Fault(Acicatena Fault by Azzaro et al., 2012). The Timpe Fault system ischaracterised by geological mean vertical slip rates ranging from 1 to2 mm/y (Monaco et al., 1997) and geodetic rates of about 10 mm/y(Bonforte et al., 2011).

On the eastern flank of Mt. Etna, several NW-SE oriented dextralfault segments, namely the S. Tecla, the Fiandaca and the TrecastagniFaults (Azzaro, 1999; Monaco et al., 2010) are present and correspondto PSF3, PSF4 and PSF6, respectively. In detail, the Fiandaca and theS. Tecla Faults, both characterised by an oblique (dextral) motion, arelinked, at their southern termination, to NNE-SSW striking normalfault segments of the Timpe Fault system, represented by the Nizzetiand the Acireale faults (Monaco et al., 1997), respectively. PSF5 is the ex-pression of the joining of the south-eastern termination of the FiandacaFault with the Nizzeti Fault. The Trecastagni Fault (Bonforte et al.,2011, 2012b; Gambino et al., 2011), showing a sharp east-facing scarp,joins, to the south, the San Gregorio-Acitrezza Fault.

Westwards, an arc-shaped extensional belt composed of normalfaults, extensional fractures and eruptive fissures, including the NERift, the summit craters and the S-Rift (Kieffer, 1975), affects themodernstratovolcano (Monaco et al., 2005) (Fig. 1). This major volcano-tectonicfeature is confined, to the north by the left-lateral E-W toWNW-ESE ori-ented Pernicana Fault (PSF1), and to the south it is confined by the PSF7that includes and joins the Mascalucia-Tremestieri and S. Gregorio-Acitrezza Faults as belonging to a unique lineament, and by the PSF8that corresponds to the NW-SE oriented transform zone cutting throughthe southern flank of Mount Etna (Catalano et al., 2004), detected lateralso by soil gas surveys and named “Belpasso-Ognina Fault” by Bonforteet al. (2012a).

A further impressive active fault segment is represented by theN-S oriented, east-dipping, Ragalna Fault System (Neri et al., 2007),located along the south-western border of the volcano to the westof the Nicolosi-Tremestieri Fault. This tectonic feature, corresponding tothe longest of three intersecting segments forming PSF9, is characterisedby dextral transtensive kinematics, downthrowing the eastern side(Bonforte and Puglisi, 2006; Monaco et al., 2010). The applicationof the Permanent Scatterers Technique on Mount Etna has alsoevidenced the occurrence of the PSF10. This NE-SW oriented featureaffects the western slope of Mount Etna, not corresponding to anyfault recognised on the field.

3. Methodology

In order to verify the activity of the PFSs during the 2003–2010 pe-riod and to detect new interferometric features, we considered a hugenumber of interferograms, evaluating the visibility of a significantmotion on each image for each structure. We analysed the DInSARinterferograms generated by the Istituto Nazionale di Geofisica eVulcanologia (INGV) – Osservatorio Etneo from SAR images acquiredby the ENVISAT satellite on Mount Etna from 2003 to 2010 in ascendingorbits and from 2006 to 2010 in descending orbits. In particular weprocessed 1277 ascending interferograms and 272 descending onesencompassing different time spans (from 35 days to several years). Inthis way wewere able to detect the PSFs short- and long-term deforma-tion styles of the PSFs that were compared with the average 1995–2000long-period behaviour, described by Bonforte et al. (2011) using ERS-1/2data.

The presence of movement of the PSFs has been assessed througha visual analysis of the DInSAR interferograms, aimed at detecting linearinterruptions and jumps in the ground deformation field along PSFs. A“visibility index” has been assigned to each PSF, in the range from 0(no visible linear interruption or jump) to 3 (extremely visible), througha value 1 (slightly visible) and 2 (visible).

The estimation of the visibility index could be partly biased by thequality of the acquired SAR images. In order to verify this hypothesis,we performed an analysis of the correlation between the visibility indexand the coherence, calculated on all interferograms (more than 1500images). In detail, we defined a box area including three PSFs locatedon the southern flank of the volcano, chosen in order to avoid seasonaleffects due to the snow coverage; we extracted the mean coherencevalue relevant to this box area;we plotted the calculated average coher-ence of each interferogram versus the respective visibility index of eachfault falling into the analysed area (Fig. 3). No correlation was found,and the results indicate that the attribution of the visibility index,even if subjective, is not biased by the coherence and then by the qualityof the interferogram.

The results of the analysis of the complete set of the interfero-grams have been summarised in correlation charts (Fig. 4), in whichthe trend of the visibility index vs. the time interval of each PSF hasbeen shown, in order to recognise the evolution of their visibility duringthe time.

4. Results

The analysis of the complete set of the interferograms has evidencedtwo distinct families of PSFs that are characterised by different behav-iours. The first family includes features showing positive slopes of thelinear regressions, then visible during the entire examined time-span,with a visibility index increasingwith the time interval. A second familygroups the PSFs that show an opposite trend, as they are characterisedby negative linear regressions (Fig. 4) due to the progressive decreasingof the visibility index as the analysed time interval extends, makingthese features visible only on some short-term interferograms.

Fig. 3. Analysis of the correlation between themean coherence value calculated on all interferograms and the visibility index. On themap, the brown box defines the test area chosen. Onthe right side, the plots of the coherence vs. visibility index for the three structures falling into the box area are reported.

132 S. Catalano et al. / Geomorphology 198 (2013) 128–137

4.1. Positive regression PSFs

The first family of positive regression PSFs comprises tectonic struc-tures that bound the arc-shaped extensional belt of the eastern flank ofMt. Etna and those characterising the western portion of the volcano.This group is composed of structures that are characterised by variabledisplacement rates with a prevalent strike-slip component of motion,showing the attitude to accumulate ground deformation through time.In this family, the most impressive feature is represented by PSF1, com-prising the NE-Rift and the Pernicana Fault (Fig. 2). The interferogramsshow that the permanent visibility of this structure is accompanied bythe progressive accumulation of a continuous displacement, highlightedby the increase of the number of interferometric fringes along the fault,due to the tectonic subsidence and/or eastward relative motion ofthe southern block of the fault. This evidence is consistent with theleft-lateral transtensive kinematics that has been reported for thisdiscontinuity by previous studies (Azzaro, 1999; Palano et al., 2006;Bonforte et al., 2007; Guglielmino et al., 2011). The cumulative displace-ment during the 7-year-long time interval examined has been estimat-ed at about 7 interferometric fringes (Fig. 5a), constraining a rate ofmotion of about 28 mm/y along the Line Of Sight (LOS). This value isconsistent with the long-term mean rates reported in the literature onthe basis of both field measurements and GPS data (Bonforte et al.,2007).

Along the southern slope of Mt. Etna, the first group of PSFs is rep-resented by the arc-shaped PSF7, which comprises the NW-orientedsegments of the Mascalucia-Tremestieri Fault and the E-W trendingS. Gregorio-Acitrezza Fault, and PSF8. During the examined time-interval, the permanent visibility of the PSF7 is accompanied by theaccumulation of a tectonic displacement of about 1 interferometricfringe, which would constrain a motion along the LOS of about 28 mmin 7 years (Fig. 5b), at rate of about 4 mm/y. Such an estimate is in per-fect agreement with the rate estimated for the period from 1995 to2000 by Bonforte et al. (2011) (3–5 mm/y) and slightly lower thanthe determination for the period from 1996 to 1998 by Froger et al.(2001) (6 mm/y).

Along PSF8, coinciding with the southern boundary of the trans-form zone reconstructed by Catalano et al. (2004), a rate of motion

of about 2 mm/y along the LOS can be estimated from the analysis of in-terferograms related to both short and long time period (Fig. 5c); thisvalue is consistent with the estimate for the period from 1995 to 2000by Bonforte et al. (2011) (2–3 mm/y along the LOS,) and again lowerthan the determination for the period from 1996 to 1998 by Froger etal. (2001) (4 mm/y).

In the western sector of Mt. Etna, PSF9, coinciding with the N-Soriented Ragalna Fault, is visible throughout the studied time-span.The interferogram (Fig. 5d) covering the entire examined period con-strains a motion-rate of about 3 mm/y along the LOS. This value iscomparably lower than both the rates calculated by Bonforte et al.(2011) (5 mm/y) for the period from 1995 to 2000, and by Frogeret al. (2001) (6 mm/y) for the period from 1996 to 1998. PSF9 linkstwo further NE-directed features, PSF9a and PSF9b whose behaviour,comparable to that of the first group, is actually based on few occur-rences of visibility that prevent us to unequivocally determine thelong-term kinematics and rate of deformation.

Another impressive feature belonging to the first group, PSF10, is lo-cated in the western flank of the volcano, to the northwest of RagalnaFault. Even if this element is not linked to any surface tectonic structure,it was observed during the whole studied time interval, with a visibilitythat increases as the time interval between the acquisition of mas-ter and slave images grows. The rate of motion, measured alongPSF10, was estimated at about 3 mm/y along the LOS (Fig. 5e).This is in perfect agreement with the calculation by Bonforte et al.(2011) who firstly detected the feature, analysing the period from1995 to 2000.

A new Permanent Scatterer Feature, here designed as PSF11(Figs. 2b, 5f), has been firstly detected through the analysis of theDInSAR interferograms. This new feature, not previously reported inthe literature, is oriented ENE-WSW running parallel to PSF10 andabout 4 km to the north of it. PSF11 shows an almost constant visibilityalways accompanying the occurrence of PSF10, so that we can include itwithin the first family of PSFs. Like PSF10, although it is visible duringthe entire period of study, PSF11 has not been related to any structureon the field, suggesting that further studies and field surveys are neces-sary to better define the geometry and kinematics of these new two tec-tonic features.

Fig. 4. Visibility index vs. time interval correlation charts.

133S. Catalano et al. / Geomorphology 198 (2013) 128–137

4.2. Negative regression PSFs

The negative regression family involves the PSFs corresponding todiscrete segments of the NNW-oriented fault segments of the TimpeFault system (PSF2 and PSF5) associated with the NW-trending faults(PSF4, PSF6) of the lower eastern flank of the volcano. This secondgroup is composed of faults that are usually characterised by a prom-inent dip-slip component of motion. The traces of the PSFs on the in-terferograms are longer than the respective mapped fault segments,whose extent has been based on the recognition of well-definedfault-scarps or fracture zones on the ground. We performed a detailedanalysis of PSF6 (Trecastagni Fault) that has been characterised by an

almost continuous and mainly aseismic creep, well evidenced by thedisplacements of several artefacts across the fault line (Gambino etal., 2011).

In contrast to this ground deformation style, the analysis of the in-terferograms revealed for this fault a period of visibility summarisedin Fig. 6, which refers to the interferograms that derive from the com-bination of 35 different master images with several different slaves,distributed through the entire studied time-interval. For each masterimage, the time interval during which the feature is not visible onintereferograms, the moment of the first appearance and, finally, thefollowing period of visibility are indicated. Fig. 6 shows that, for each se-lected image, the feature appears on interferograms only when it is

Fig. 5. Interferograms showing the PSFs belonging to the “positive regression PSFs” (blue lines in figure). This group is composed of structures that are characterised by variabledisplacement rates, showing the attitude to accumulate ground deformation through time (progressive behaviour).

134 S. Catalano et al. / Geomorphology 198 (2013) 128–137

combined with another acquired (at least) one year later and that thetime-interval of visibility of the feature is almost constant in duration(3 years), after the first appearance. The figure also shows that the im-ages acquired in sequence from the first one (considered the masterimage) to the moment of the first appearance, if combined with eachother, generate interferograms in which the feature is totally absent(grey boxes in Fig. 6). They simulate periods of quiescence of the struc-ture, separating episodic appearances, which could suggest repeated

reactivations of the fault. The analyses of the series of interferogramsevidenced, at least, 5 distinct appearances of the feature.

5. Discussions and conclusions

Analysis of the interferograms deriving from the combination ofSAR images of Mount Etna, for the period from 2003 to 2010 hasallowed a short- to long-term characterization of the known PSFs. In

Fig. 6. (a) Time spans covered by the interferograms marking the absence (dashed lines), the moment of the first appearance and the following period of visibility (continuouslines) of the PSF6 (dots refers to the different master SAR images; grey boxes indicate the time period in which the feature is absent); (b) Monthly rainfalls at Pedara meteorologicalstation for the period from 2003 to 2010 (data provided by SIAS – Servizio Informativo Agrometeorologico Siciliano).

135S. Catalano et al. / Geomorphology 198 (2013) 128–137

addition, a new NE-SW oriented PSF, namely PSF 11, has been detectedon the north-western flank of the volcano. The PSF11 is visible through-out the entire analysed time-interval; it is not related to any known

Fig. 7. Interferograms showing the absence of the effects of ground deformation along thelines in figure). The intermittent PSFs are visible when one of the two (tectonic vs. erosionover the other.

structure on the field, as for the neighbour PSF10, suggesting that fur-ther studies and field surveys are necessary to better define its role inthe active deformation of the volcano.

PSFs belonging to the “negative regression PSFs” with intermittent behaviour (brownal) interacting processes governing the topographic changes around the fault prevails

136 S. Catalano et al. / Geomorphology 198 (2013) 128–137

The analysed interferograms provided significant short-term pic-tures of the active deformation in the volcanic region, during the en-tire 2003–2010 time span. Two distinct families of structures havebeen recognised on the basis of their visibility vs. time. Some of thePSFs are visible throughout the examined time-span, with a visibilityindex increasingwith time. This behaviour is that expected for structuresaccumulating amounts of continuous ground deformation (progressivebehaviour). In contrast, other PSFs are characterised by episodic appear-ances, alternatingwith periodswith lower or null visibility index (Fig. 7).This behaviour is interpreted as a different ground deformation style, re-lated to a stick–slip mechanism (intermittent behaviour). The differentrate of deformation of the structures could be ruled out as the primarycause to explain the two opposite behaviours of the PSFs of MountEtna: (i) the first family groups faults that are characterised by awide range of displacement-rate, from 28 mm/y (e.g. PernicanaFault) to 4 mm/y (e.g. Nicolosi-Tremestieri Fault); (ii) the secondfamily comprises only structures with a lower displacement rate(about 1–2 mm/y; S. Leonardello Fault). However, the relation betweenthe appearance of the PSFs on the interferograms with episodic butpermanent ground deformation is inconsistent with a decrease (afterits appearance) of the visibility index to zero. The different partitioningof the vertical and lateral deformation along the structures of thetwo families may condition the meaning of their visibility on theinterferograms. The structures belonging to the two different familiescan be, in fact, discriminated also on the basis of their kinematics. It isnoteworthy that the structures with progressive behaviour (progressivePSFs) are mainly characterised by strike-slip component of motion,while the structures with intermittent behaviour (intermittent PSFs),are affected by a prevalent dip-slip component. This would imply thatthe motion along the faults belonging to the intermittent PSFs producessignificant vertical ground offsets, triggering surface mass movementsfrom the up-thrown to the down-thrown block. In the case of the struc-tures of the progressive PSFs, the visibility refers to topographic changesdue mainly to the lateral tectonic displacement. In the case of the struc-tures of the intermittent PSFs, the visibility results from topographicchanges due to the combination between the vertical tectonic displace-ments and the related amounts of differential erosion and depositionacross the fault line. We validated this relation by correlating the ap-pearances of the PSF6 with the monthly rainfalls in the area (Fig. 6b).In figure it is possible to observe that the appearances generally occurafter prolonged dry periods during which the structure can accumulatedisplacements without any significant episode of sculpturing across thefault line. The appearance of the structure is also related to an impres-sive heavy rain episode on 27th December 2006 (267 mm). This eventproduced a transient visibility of the feature, probably due to the hugemass transfer across the fault, causing significant topographic changesalong the LOS of the radar sensor. In fact, only in this case we observedan apparent opposite displacement, measured a few hundred metresacross the fault trace, imputable to a significant erosion that loweredthe top of the fault scarp (footwall) and deposited material at its base,on the hangingwall. On the contrary, the feature is absent on interfero-grams deriving from the series of images acquired during the sameseasonal wet period, when the runoff and mass transport from the up-thrown to the downthrown block probabily almost balance the groundtectonic deformation. The intermittent PSFswould, thus, be visible whenone of the two (tectonic vs. erosional) interacting processes governingthe topographic changes around the fault prevails over the other.

In conclusion, our study evidences that even if the entire set of ac-tive faults of Mount Etna accumulated significant amounts of almostcontinuous deformation during the period 2003–2010, their expres-sion on DInSAR interferograms show two distinct behaviours whichare mostly related to the kinematics of the structures. The fault linescharacterised by a prevalent strike-slip component of motion appearto accumulate displacements on a series of interferograms that coveran increasing time-span. In contrast, along the fault lines moving witha prevalent dip-slip component, the effects of mass transport from the

upthrown to the downthrown block, causing the progressive obliter-ation of the topographic offsets, strongly influenced the visibility,which is most likely governed by the seasonal variation in rainfall,rather than by episodic remobilisation of the faults.

Acknowledgments

We thank the two anonymous referees for the constructive com-ments and the precious suggestions.

This work was funded by University of Catania grants (Prof. StefanoCatalano) and by the Task D7 “Enhancement of the remote sensing lab-oratory” of the project “Extension and enhancement of the volcanic andseismic monitoring systems of Sicily”, funded by the Sicilian regionalgovernment. The ENVISAT data were provided in the frame of the ESACAT.1 5843 project.

References

Azzaro, R., 1999. Earthquake surface faulting at Mount Etna volcano (Sicily) and implica-tions for active tectonics. Journal of Geodynamics 28, 193–213.

Azzaro, R., Branca, S., Giammanco, S., Guerrieri, S., Rasà, R., Valenza, M., 1998. New ev-idence for the form and extent of the Pernicana Fault System (Mt. Etna) from struc-tural and soil-gas surveying. Journal of Volcanology and Geothermal Research 84,143–152.

Azzaro, R., Branca, S., Gwinner, K., Coltelli, M., 2012. The volcano-tectonic map of Etnavolcano, 1:100.000 scale: an integrated approach based on amorphotectonic analysisfromhigh-resolution DEM constrained by geologic, active faulting and seismotectonicdata. Bollettino della Società Geologica Italiana 131, 153–170.

Barreca, G., Bonforte, A., Neri, M., 2013. A pilot GIS DataBase of active faults of Mt. Etna(Sicily): a tool for integrated hazard evaluation. Journal of Volcanology and Geo-thermal Research 251, 170–186.

Bonaccorso, A., Bonforte, A., Currenti, G., Del Negro, C., Di Stefano, A., Greco, F., 2011.Magma storage, eruptive activity and flank instability: Inferences from ground defor-mation and gravity changes during the 1993–2000 recharging of Mt. Etna volcano.Journal of Volcanology and Geothermal Research 200, 245–254. http://dx.doi.org/10.1016/j.jvolgeores.2011.01.001.

Bonforte, A., Puglisi, G., 2006. Dynamics of the eastern flank of Mt. Etna volcano (Italy)investigated by a dense GPS network. Journal of Volcanology and Geothermal Re-search 153, 357–369.

Bonforte, A., Branca, S., Palano, M., 2007. Geometric and kinematic variations along theactive Pernicana fault: implication for the dynamics of Mount Etna NE flank (Italy).Journal of Volcanology and Geothermal Research 160, 210–222.

Bonforte, A., Bonaccorso, A., Guglielmino, F., Palano, M., Puglisi, G., 2008. Feeding sys-tem and magma storage beneath Mt. Etna as revealed by recent inflation/deflationcycles. Journal of Geophysical Research, B: Solid Earth 113 (5), B05406. http://dx.doi.org/10.1029/2007JB005334.

Bonforte, A., Guglielmino, F., Coltelli, M., Ferretti, A., Puglisi, G., 2011. Structural assess-ment of Mount Etna volcano from Permanent Scatterers analysis. Geochemistry,Geophysics, Geosystems 12, Q02002. http://dx.doi.org/10.1029/2010GC003213.

Bonforte, A., Federico, C., Giammanco, S., Guglielmino, F., Liuzzo, M., Neri, M., 2012a.Soil gases and SAR measurements reveal hidden faults on the sliding flank of Mt.Etna (Italy). Journal of Volcanology and Geothermal Research. http://dx.doi.org/10.1016/j.jvolgeores.2012.08.010.

Bonforte, A., Carnazzo, A., Gambino, S., Guglielmino, F., Obrizzo, F., Puglisi, G., 2012b. Amultidisciplinary study of an active fault crossing urban areas: The TrecastagniFault at Mt. Etna (Italy). Journal of Volcanology and Geothermal Research. http://dx.doi.org/10.1016/j.jvolgeores.2012.05.001.

Branca, S., Coltelli, M., De Beni, E., Wijbrans, J., 2007. Geological evolution of Mount Etnavolcano (Italy) from earliest products until the first central volcanism (between 500and 100 ka ago) inferred from geochronological and stratigraphic data. InternationalJournal of Earth Sciences 97, 135–152.

Catalano, S., Torrisi, S., Ferlito, C., 2004. The relationship between Late Quaternary de-formation and volcanism of Mt. Etna (eastern Sicily): new evidence from the sed-imentary substratum in the Catania region. Journal of Volcanology and GeothermalResearch 132, 311–344.

Catalano, S., De Guidi, G., Monaco, C., Tortorici, G., Tortorici, L., 2008. Active faulting andseismicity along the Siculo-Calabrian rift zone. Tectonophysics 453, 177–192.

Catalano, S., Romagnoli, G., Tortorici, G., 2010. Rift-flank deformation in SE Sicily: evi-dence for an active Mantle upwelling. Tectonophysics 468, 1–14.

Catalano, S., Torrisi, S., Tortorici, G., Romagnoli, G., 2011. Active folding along a rift flank:the Catania region case history (SE Sicily). Journal of Geodynamics 51, 53–63.

Curlander, J.C., McDonough, R.N., 1991. Synthetic Aperture Radar: Systems and SignalProcessing. John Wiley, New York.

Ferretti, A., Prati, C., Rocca, F., 2001. Permanent Scatterers in SAR interferometry. IEEETransactions on Geoscience and Remote Sensing 39, 8–20.

Froger, J.L., Merle, O., Briole, P., 2001. Active spreading and regional extension at MountEtna imaged by SAR interferometry. Earth and Planetary Science Letters 187,245–258.

Gambino, S., Bonforte, A., Carnazzo, A., Falzone, G., Ferrari, F., Ferro, A., Guglielmino, F.,Laudani, G., Maiolino, V., Puglisi, G., 2011. Displacement across the Trecastagni

137S. Catalano et al. / Geomorphology 198 (2013) 128–137

Fault (Mt. Etna) and induced seismicity: the October 2009 to January 2010 episode.Annals of Geophysics 54 (4), 414–423.

Gillot, P.Y., Kieffer, G., Romano, R., 1994. The evolution of Mount Etna in the light ofpotassium-argon dating. Acta Vulcanologica 5, 81–87.

Guglielmino, F., Bignami, C., Bonforte, A., Briole, P., Obrizzo, F., Puglisi, G., Stramondo, S.,Wegmuller, U., 2011. Analysis of satellite and in situ ground deformation data inte-grated by the SISTEM approach: the April 3, 2010 earthquake along the Pernicanafault (Mt. Etna – Italy) case study. Earth and Planetary Science Letters 312, 327–336.

Houlie, N., Briole, P., Bonforte, A., Puglisi, G., 2006. Large scale ground deformation ofEtna observed by GPS between 1994 and 2001. Geophysical Research Letters 33(2), L02309. http://dx.doi.org/10.1029/2005GL024414.

Kieffer, G., 1975. Sur l'existence d'une “rift-zone” à l'Etna. Comptes Rendus de l'Academie des Sciences, D 280, 263–266.

Massonnet, D., Feigl, K.L., 1998. Radar interferometry and its application to changes inthe Earth's surface. Review of Geophysics 36, 441–500.

Monaco, C., Tapponier, P., Tortorici, L., Gillot, P.Y., 1997. Late Quaternary slip rates on theAcireale-Piedimonte normal faults and tectonic origin of Mt. Etna (Sicily). Earth andPlanetary Science Letters 147, 125–139.

Monaco, C., Catalano, S., Cocina, O., De Guidi, G., Ferlito, C., Gresta, S., Musumeci, C.,Tortorici, L., 2005. Tectonic control on the eruptive dynamics at Mt. Etna Volcano

(Sicily) during the 2001–2003 eruptions. Journal of Volcanology and GeothermalResearch 144, 211–233.

Monaco, C., De Guidi, G., Ferlito, C., 2010. The Morphotectonic map of Mt. Etna. ItalianJournal of Geoscience 129 (3), 408–428.

Neri, M., Guglielmino, F., Rust, D., 2007. Flank instability of Mount Etna: Radon, radarinterferometry and geodetic data from the southwestern boundary of the unstablesector. Journal of Geophysical Research 112, B04410. http://dx.doi.org/10.1029/2006jB0047.

Palano, M., Aloisi, M., Amore, M., Bonforte, A., Calvagna, F., Cantarero, M., Consoli, O.,Guglielmino, F., Mattia, M., Puglisi, B., Puglisi, G., 2006. Kinematics and strain analysesof the eastern segment of the Pernicana Fault (Mt. Etna, Italy) derived from geodetictechniques (1997–2005). Annals of Geophysics 49 (4/5) (August/October 2006).

Rasà, R., Azzaro, R., Leonardi, O., 1996. Aseismic creep on faults and flank instability atMount Etna volcano, Sicily. Volcanic hazards and their mitigation. Geological Soci-ety, London, Engineering Geology Special Publications 15, 79–95.

Serpelloni, E., Vannucci, G., Pondrelli, S., Argnani, A., Casula, G., Anzidei, M., Baldi, P.,Gasperini, I., 2007. Kinematics of the Western Africa-Eurasia plate boundary fromfocal mechanisms and GPS data. Geophysical Journal International 169, 1180–1200.http://dx.doi.org/10.1111/j.1365-246X.2007.03367.x.