Propagation of an inherited strike-slip fault through … et al., 2009... · mazioni a carico di un...

ABSTRACT

We investigated quantitatively the propagation of a reactivatedstrike-slip fault through a sedimentary cover. To this end we pre-pared five simplified analogue models that reproduce a chain withits frontal allochtonous wedge overrunning the foreland. The fore-land/chain deformation follows the reactivation of an inheritedstrike-slip fault cutting the foreland domain. The observation andquantification of the effects of this reactivation, in particular on theorogenic wedge front, provide new insight on the evolution of thistype of tectonic setting. We placed special emphasis on quantifyingthe structural features observed in the models to (1) interpret thekinematics of the reactivated shear zone, and (2) put forwardhypotheses on areas indirectly affected by the reactivated fault. Theinterpretation of the models was based on an integrated analysis ofsurface and subsurface data. The results show that the geologicalsetting is strongly influenced by the presence of a reactivated pre-existing lineament, that ultimately controls the development andpattern of newly-formed faults. Finally, we present and discuss twonatural examples (in Italy Molise-Gondola shear zone, SouthernApennines, and Scicli-Ragusa line, Sicily) in view of the modelingresults.

KEY WORDS: Fault reactivation, foreland, orogenic wedge,sandbox models, quantitative analysis.

RIASSUNTO

Propagazione di una faglia trascorrente ereditata attraversoun sistema avampaese/catena: aspetti quantitativi e applicazionisulla base di modelli analogici.

Nel presente lavoro è stata investigata quantitativamente lapropagazione di una zona di taglio regionale, trascorrente, eredita-ta che, riattivandosi, fa risentire i propri effetti su una coperturasedimentaria. Tale copertura è data dalla parte frontale di una ca-tena caratterizzata da unità alloctone (cuneo orogenico), marcata-mente sovrascorse sul relativo avampaese in cui è presente la zonadi taglio stessa. Per questo motivo, sono stati realizzati cinque modelli analogici che riproducono in modo semplificato le defor-mazioni a carico di un sistema costituito da avampaese-cuneo oro-genico-catena, conseguenti alla riattivazione di una faglia tra-scorrente ereditata sita all’interno dell’avampaese e ortogonale alfronte della catena. Le osservazioni e le analisi quantitative degli

effetti della riattivazione sul cuneo orogenico e sulla catena, possi-bili grazie alla grande quantità di dati facilmente ricavabili dai mo-delli, forniscono in particolare ulteriori e nuovi spunti sulla cono-scenza della possibile evoluzione di questi sistemi e contestistrutturali. Particolare attenzione è stata dedicata agli aspettiquantitativi di elementi strutturali osservati nei modelli per (1) in-terpretare correttamente la cinematica del lineamento riattivato e(2) formulare osservazioni e ipotesi sui domini che non ospitanodirettamente il lineamento riattivato (cuneo orogenico e catena).L’interpretazione dei modelli si è basata sull’osservazione e l’analisicongiunta dei dati raccolti in superficie e in profondità (superficiedel modello e sezioni del modello stesso). I risultati mostrano cheil contesto geologico è fortemente influenzato dalla presenza di unlineamento preesistente che si riattiva e che controlla lo sviluppodelle faglie di nuova formazione. Al termine del lavoro, vengonoproposti e discussi due esempi di casi naturali in Italia (Molise-Gondola shear zone, Appennino Meridionale e linea Scicli-Ragusa,Sicilia) alla luce dei risultati dei modelli analogici qui presentati.

TERMINI CHIAVE: Riattivazione, avampaese, cuneo orogeni-co, modelli analogici, analisi quantitativa.

1. INTRODUCTION

This paper follows and integrates a previous study (DI

BUCCI et alii, 2006) where a set of analogue models wasstudied to clarify the seismotectonic setting of a portionof the Southern Apennines chain and related foreland. Inthe previous paper, we thoroughly analysed the seismo-tectonic implications of this set of models. However, aquantitative structural investigation and a detailed analy-sis of the propagation of the deformation through themodels (i.e. through different structural domains) werestill missing. For this reason, we carried out the quantita-tive study here presented, that has to be considered as acomplement of the previous work.

The paper is organised in three parts. The first one isa short description of the analogue models presented byDI BUCCI et alii (2006) and resumes the main resultsobtained by these investigators (paragraphs 1, 2 and 3).The second part is completely new, and is focussed on thequantitative analysis of the models and of their kinemat-ics and evolution through time (paragraphs 4, 5, 6 and 7).At the end of the paper, the third part corresponds to anappendix devoted to two real geological cases, showingsimilar structural conditions, whose interpretation couldbenefit from insights provided by the analogue modelsdescribed.

A reactivated strike-slip fault may occur in varioustectonic settings including foredeeep basins, volcanic

Boll.Soc.Geol.It. (Ital.J.Geosci.), Vol. 128, No. 1 (2009), pp. 000-000, 12 figs., 1 tab.

Propagation of an inherited strike-slip fault through a foreland-chain system:

quantitative aspects from analogue modeling and applications

GIOVANNI TOSCANI (*), DANIELA DI BUCCI (**), ANTONIO RAVAGLIA (*), (***), SILVIO SENO (*), UMBERTO FRACASSI (****) & GIANLUCA VALENSISE (****)

(*) Dipartimento di Scienze della Terra, Università di Pa-via, Via Ferrata, 1 - 27100 Pavia, Italy. Tel. +39 0382 985857 - Fax+39 0382 985890 - e-mail: [email protected]

(**) Dipartimento della Protezione Civile, Via Vitorchiano, 4- 00189 Roma, Italy.

(***) PetroSA, 151 Franns Conradie Drive, Parow 7500, Repu-blic of South Africa.

(****) Istituto Nazionale di Geofisica e Vulcanologia, Via diVigna Murata, 605 - 00143 Roma, Italy.

SGI 12/08-227

Queste bozze, corrette e accompagnate dall’al-legato preventivo firmato e dal buono d’ordine,debbono essere restituite immediatamente allaSegreteria della Società Geologica Italianac/o Dipartimento di Scienze della TerraPiazzale Aldo Moro, 5 – 00185 ROMA

domains and chains (a chain is here intended as a stackof thrust sheets with its allochtonous frontal wedge, a fewkilometers thick, unaffected by regional vertical strike-slip faults coinciding with those existing underneath, inthe buried foreland). Positively identifying a buried faultthrough the structural analysis of the overlying chain isespecially important in active tectonics, as it may allowblind or simply unidentified seismogenic structures to bedetected and taken into account in seismic hazard andscenario studies. For this reason, we developed and ana-lyzed a set of sandbox models, aimed at (i) investigatinghow strike-slip motion along a pre-existing zone of weak-ness, exposed or buried, propagates toward the surfaceand affects the cover belonging to a chain-foreland sys-tem; and (ii) characterizing the propagation of deforma-tion from this inherited structure as a function of dis-placement.

When propagating through a chain and its «allochto-nous» frontal wedge, a buried, reactivated, large strike-slip fault is expected to exhibit the same behavior as anew fault growing in a brittle environment (fig. 1). Thepropagation is preceded by pervasive brittle damage ofthe host rock. Due to the perturbation of the stress field atthe tips of the fault, strain accumulates in these regionsas the fault grows (MANDL, 2000; and references therein).Strain at the tips of a fault has been traditionallydescribed in terms of modes of crack propagation(SCHOLZ, 1990; and references therein), where Mode-I isthe tensile or opening mode (crack wall displacementsnormal to the crack), and Mode-II (displacements in theplane of the crack and normal to the crack edge) and -III(displacements in the plane of the crack and parallel tothe edge) are the shear modes. For instance, in case of avertical strike-slip fault, the Mode-II is expected at thevertical tips of the fault, the Mode-III at the horizontaltips, and a combination of them along the oblique tips(KIM et alii, 2004; KIM & SANDERSON, 2006; fig. 1).

2. EXPERIMENTAL SET-UP

We carried out a set of models, deformed in a naturalgravity field, aimed at reproducing the crustal volumeformed by a foreland hosting an inherited discontinuity,

namely a sub-vertical strike-slip fault, and a chain with itsfrontal wedge overrunning the faulted foreland (fig. 2). Allmodels were scaled 1:200,000 (0.5 cm = 1 km in nature).In scaling the model setup, we took into account geomet-ric, kinematic and dynamic similarity relationships (HUB-BERT, 1937; RAMBERG, 1981). The relation Cp/rp = 5×106

Cm/rm links natural and experimental conditions. C is thecohesion, r is the density in nature (subscript p) andmodels (subscript m), respectively. Knowing that rm is1500 kgm-3 and rp is 2800 kgm-3, the cohesion in themodel (~10 Pa) corresponds to ~100MPa in nature. Thefrictional properties of granular materials used for themodels have been measured using a Casagrande shearbox (CASAGRANDE, 1932). The cohesion has been derivedindirectly by extrapolating the value from severalCasagrande shear box tests.

We used two different granular materials: dry sandand glass microbeads. The angles of internal friction (f)are equal both in the models and in nature as the anglesmeasured on the analogue materials (the angle of a sta-ble cone of sand/microbeads) are comparable to thosedetermined experimentally for upper crustal rocks(BYERLEE, 1978). The sand has f = 33° and grain size inthe range 100-300 mm. Glass microbeads have f = 24°and grain size in the range 300-400 mm. Their low inter-nal friction is due to high sphericity and rounding(SCHELLART, 2000) so that glass microbeads enable lowbasal friction detachment and inter-strata slip (SASSI etalii, 1993; TURRINI et alii, 2001). The basal detachmenthas f = 32°. For the quantitative description of experi-mental apparatus and all materials from a rheologicalpoint of view we adopted the values described bySCHREURS et alii (2006).

The apparatus was provided with a baseplate fault,which extended for the whole length of the models andaccommodated a simple right-lateral shear. The basal dis-placement varied for each model from 0.5 cm to 8.0 cm.

We prepared five independent models. The first (SS02;fig. 3) reproduces a typical wrench zone as widelydescribed in literature (among many others: WILCOX etalii, 1973; CHRISTIE-BLICK & BIDDLE, 1985; HARDING,1985; NAYLOR et alii, 1986; SYLVESTER, 1988; RICHARD &COBBOLD, 1990; RICHARD et alii, 1995; MANDL, 2000;NIEUWLAND & NIJMAN, 2001; LE GUERROUÉ & COBBOLD,

2 G. TOSCANI ET ALII

Fig. 1 - Block-diagram showing the modes of crack propagation atthe tip lines of a right-lateral strike-slip fault of the foreland throughthe chain.– Blocco-diagramma che mostra la propagazione della deformazione ai li-miti di una faglia trascorrente destra dell’avampaese attraverso la catena.

Fig. 2 - Experimental set-up. Two fixed sidewalls confine the sand,parallel to the strike-slip motion, whereas the model is open on theother two sides. The three regional-scale structural domains dis-cussed in the text are indicated as A, B, C.– Set-up dei modelli. Due pareti laterali, parallele al movimento dellafaglia, confinano la sabbia, mentre sugli altri due lati il modello non èconfinato. I tre domini strutturali di cui si tratta nel testo sono indicaticon A, B, C.

2006). It has a constant thickness of 10 cm, its sand vol-ume has constant physical and rheological characteristics(neither layers of glass microbeads, nor cuts), and the totaldisplacement applied to its baseplate fault is 8.0 cm. Thismodel was used as a reference for four additional models(SS03 to SS06, shown in map view in fig. 4), all specifi-cally designed to reproduce the foreland-orogenic wedge-chain system. We added to these models a layer of glassmicrobeads at the interface between the buried forelandand the orogenic wedge, aimed at simulating the highlytectonized zone in correspondence of the basal thrust sep-arating the chain from its foreland (fig. 2). The chain-sideof the model is slightly thicker (11 cm) than the foreland-side (10 cm) to account for topography. The models SS03-SS06 were subjected to progressively smaller displace-ments: 8.0, 5.5, 3.0 and 0.5 cm, respectively.

Within the entire foreland domain (both the exposedregion and the portion lying below the wedge) a 5 mm-thick layer of glass microbeads simulates the presence ofa thick layer (~1,000 m) of Triassic anhydrite-dolomiterocks at the bottom of the Apulia platform. The forelanddomain is intersected by a vertical discontinuity perpen-dicular to the wedge front, obtained through a cut thatreorganizes the grain distribution (SASSI et alii, 1993;VIOLA et alii, 2004) and creates a preferential slip surface.This discontinuity starts from the baseplate fault and hasa vertical attitude. Conversely, no discontinuities wereintroduced in the chain-side portion of the model and inthe wedge itself.

The model set-up aims at reproducing the three mainregional-scale domains which together describe the tec-tonic setting we intended to study:

– domain A - the foreland;– domain B - the frontal wedge (including the fore-

deep deposits);– domain C - the chain.

In the following these domains will be referred tosimply as A, B, C.

PROPAGATION OF AN INHERITED STRIKE-SLIP FAULT THROUGH A FORELAND-CHAIN SYSTEM 3

Fig. 3 - Interpreted plan-views of the final step of reference modelSS02 (top) and model SS03 (bottom). Reference vertical lines arespaced ca. 5.5 cm. The horizontal hatched line is the baseplate fault.Final displacement was D = 8.0 cm for both models. – Interpretazione della visione in mappa della fase deformativa finaledei modelli SS02 (in alto) e del modello SS03 (in basso). Le linee verti-cali di riferimento sono spaziate di circa 5.5 cm. La linea orizzontaletratteggiata rappresenta ed indica la faglia di base. Per entrambi i mo-delli il rigetto finale è pari a 8 cm.

Fig. 4 - Plan-view of the SS06, SS05, SS04 and SS03 models (fromtop to bottom) at the end of the deformation. The black line indi-cates the front of the orogenic-wedge; the hatched white line is theprojection of its bottom-end on the model surface.– Mappa dei modelli SS06, SS05, SS04 e SS03 (dall’alto in basso) altermine della deformazione. La linea nera indica il fronte del cuneoorogenico; la linea bianca tratteggiata è la proiezione sulla superficiedella sua terminazione alla base dell’esperimento.

3. COMPARISON WITH PREVIOUS SIMILAR EXPERIMENTS(STRIKE-SLIP, REACTIVATION)

AND LIMITATIONS OF THE MODELS

Many workers investigated strike-slip mechanismsand fault reactivation in different geological settingsusing analogue models and demonstrated how pre-exist-ing lineaments influence the pattern of strike-slip faults inthe overlying deposits (NAYLOR et alii, 1986; MANDL,1988; RICHARD et alii, 1995). The «novelty» of our experi-ments is that we modeled and analyzed together three dif-ferent structural domains that interact during the defor-mation. This allowed us to bridge the gap between anumber of natural examples and large part of the existingliterature, which is generally focused on single structuraldomains.

Modeling shear zones at different scales has shownthat the rheological properties of the materials used andtheir thickness have the greatest impact on the finalwidth of the shear zone and the length of the associatedfaults (TCHALENKO, 1970). Different stratigraphies havebeen tested (brittle and brittle-viscous) to study the geo-metrical pattern of faults, both in map and in cross sec-tion (RICHARD et alii, 1989; RICHARD & COBBOLD, 1990;RICHARD & KRANTZ, 1991). All these studies confirmedthat the fault geometry and kinematics are strongly influ-enced by rheology and that flower structures or palm treefaults may develop even in the absence of transpressivedeformation. Fault dip within the flower structuredecreases even further whenever a compressional stresscomponent is introduced. The width and complexity ofthe shear zones vary as displacement increases.

As in any modeling technique, sandbox analoguemodels have several limitations that must be taken intoaccount while reproducing geological settings and analyz-ing results. In some cases, the scale of the model is a limi-tation itself; the presence of natural thin viscous layerscannot be simulated, as their modeled thickness would betoo small to be reproduced in a sand-box apparatus. Forexample, in our models the rheological structure of thenatural thrust wedge and of the chain is reproduced by asystem composed only of brittle rocks. The presence ofinternal layers, faults, fractures, zones of weakness, etc.,all features that occur everywhere in natural cases, ishence neglected, so that these parts of the models areisotropic. In addition, faulting in the upper crustresponds to variables like pore-fluid pressure, heat flow,presence of overpressured shales, and flexural and isosta-tic effects that are difficult to simulate in a sand-boxmodel. Extreme folding is also impossible to reproducewith sand and microbeads only. Finally, an undesiredeffect that often occurs in sand-box models is dilation,that may cause a volume increase (up to 5% and beyond)depending on grain size, sphericity, rounding and initialgathering conditions of the material (SCHOPFER &STREYER, 2001). Nevertheless, since our work aims atsimulating and studying the overall behavior of the com-plex system composed by the aforementioned tectonicdomains, we are confident that these limitations do notprevent large-scale quantitative analyses as those pre-sented in the following paragraphs. This is suggested andsupported by the good geometric similarities between theanalogue models and the natural cases that we will dis-cuss at the end of the paper for a comparison.


Fig. 5 - Interpreted line drawing of fig. 4. Different structural domains are represented from left to right with different grey colours (A=fore-land=heavy grey; B=wedge=medium grey; C=chain=light grey).– Line-drawing di quanto mostrato in fig. 4. I differenti domini strutturali sono rappresentati da sinistra verso destra con differenti gradazioni digrigio (A=foreland=grigio scuro; B=cuneo orogenico=grigio intermedio; C=catena=grigio chiaro).

2. EXPERIMENTAL RESULTS

4.1. MODEL EVOLUTION

Surface structures can be observed simply by lookingat the experiments as they evolve (fig. 5). On the contrary,internal structures can be analyzed only by cutting (thatis to say, destroying) the models, unless non-invasivemeans of analysis (e.g., X-Ray tomography) are available.A detailed analysis of the model evolution can be found inDI BUCCI et alii (2006) where the reference model and theSS03 model are compared step by step and where allfaults are distinguished on the base of their activity; forthis reason, in this paragraph, we skip this descriptionand analysis in order to focus the attention on the mapsof the four models, representing the «bench marks» forthe quantitative analysis here presented. Nevertheless, acontinuous analysis of the model evolution has been car-ried out (as the measurements have been taken at every0.5 cm of deformation), even if it is not shown usingevery step of the models but only the maps of the mainfour models.

As mentioned earlier, we prepared a model withouttopographic variations and discontinuities (no pre-cut,no layers of glass microbeads) as a «reference» to whichto compare the other models of this study. fig. 3 showsthe fault pattern at the surface of the reference modelSS02 (top) and of model SS03 (bottom); for both theseexperiments the displacement of the baseplate fault was8.0 cm. In SS03 new faults form only in domains B andC. In domain A, the pre-existing fault acts as a preferen-tial slip surface accommodating almost all the displace-ment; in fact, no newly formed faults will ever show upin domain A.

The model sequence from SS06 to SS03 documentsthe progressive deformation of the same shear zone. Asthe basal displacement increases from 0.5 cm up to 3.0cm (models SS06 and SS05, respectively) a number ofstructures form near the front of the orogenic wedge(domain B) and in particular in the northern (i.e. reced-ing) block exclusively (fig. 5). At 3.0 cm of displacement,the model (SS05, figs. 4 and 5) exhibits a certain concen-tration of active faults in its central part and an increas-ing development of faults deviating from the E-W strike.In foreland domain A, the only active fault is the pre-existing one, whereas in the orogenic wedge domain Bfaults are still present only in its northern portion. At 4.5

cm of basal displacement, active faults are detectable overthe entire length of the models. These faults strike quiteclose to the strike of the baseplate fault and tend to belocated in a central limited band, parallel to the sheardirection. The final steps corresponding to 5.5 cm and 8.0cm of displacement (models SS04 and SS03) show theevolution of the experiment when a «steady state» isreached and overrun, i.e. when the northern and southernblocks are completely separated by faults and move inde-pendently. In these last steps of deformation, the onlyactive faults are the central ones. We do not observe anynew active faults between 5.5 and 8.0 cm of basal dis-placement, but only an increase of displacement onalready existing faults that are suitably oriented withrespect to the shear direction.

The kinematics of the models shows how the initialset-up influences the development of brittle structures. Inall models the presence of a pre-existing cut prevents theformation of new faults in domain A. In domain B (oro-genic wedge) all the brittle structures develop asymmetri-cally turning northward from the E-W strike. Finally,structural features of domain C are similar to those seenin models reproducing purely strike-slip displacement(see SS02, fig. 3).

4.2. FINAL MODEL GEOMETRIES

The deformation seen at the surface of model SS03(no newly-formed faults in foreland domain A, asymmet-ric fault distribution in wedge domain B, widespreadfaulting in chain domain C; figs. 3, 4 and 5) has its coun-terpart in the vertical sections of all the models, as exem-plified by model SS05 (figs. 6 and 7). The sections used tocreate the 3D view are perpendicular to the baseplatefault and typify the three domains A, B and C of theexperiments.

In domain A (S 66; figs 6 and 7), we observe the com-plete absence of new faults, confirming that the inheritedfault takes up the entire imposed displacement.

In domain B (S 54), incipient faulting is evident sincethe very early steps of deformation (model SS06; 0.5 cmof displacement, fig. 5). As deformation increases, thesenew faults also grow in number and displacement. For0.5 cm of basal displacement, the only existing faulthardly reaches the topographic surface with almost noapparent vertical throw. On the contrary, faults of the


Fig. 6 - Sketched 3D reconstruction of the central partof the model SS05 (D=3.0 cm), see fig. 5, obtainedusing sections perpendicular to the baseplate fault.Black lines on sections represent interpreted faults.– Ricostruzione schematica 3D della parte centrale delmodello SS05 (D=3.0 cm), vedi fig. 5, ottenuta utilizzan-do sezioni perpendicolari alla faglia di base. Le linee neresulle sezioni rappresentano le faglie interpretate.

same domain B cause a detectable topographic effect(partially visible in fig. 6, S 42) in the other models (SS05,SS04 and SS03). This implies a dip-slip component ofmotion on the strike-slip faults. This component is eithernormal or reverse, and sometimes a transition fromreverse to normal is seen along strike of the same faultplane. Vertical sections of domains A and B also showboth the layer of glass microbeads that separates the oro-genic wedge from the underlying foreland (frontal thrust)and that in the foreland stratigraphic sequence. Theselayers had no effects on the inception and development ofthe faults, regardless of the amount of displacement. Thisis due to the set-up of the models (existence of the pre-cutin domain A, pure strike-slip movement of the twoblocks), for which the layers of glass microbeads did notundergo compression but only strike-slip deformation.Moreover, no faults propagate through the layers ofmicrobeads in our models; therefore, the differences inrheology they induce do not affect the shape of the newly-formed faults. Instead, all newly-formed faults originatefrom the pre-existing vertical fault that exists in theburied foreland and affects the eastern portion of themodel (the northern block in map-view).

Finally, in domain C (S 34 and S 26; figs. 6 and 7)faults splay up from the baseplate fault and affect quitesymmetrically the overburden. They are mainly strike-slipbut sometimes exhibit a reverse component producing avertical throw that varies from model to model. Topo-graphic relief is largest in this domain. Both the reversecomponent of slip and the topographic relief have beenrecently interpreted by LE GUERROUÉ & COBBOLD (2006)as due to rock dilation, which simulates transpressive fea-tures in spite of the simple shear applied.

5. QUANTITATIVE ANALYSIS

We carried out a quantitative analysis by measuringsystematically the actual displacements seen along thedifferent faults of models SS03 to SS06 and comparingthem with the displacement applied to the baseplate fault.fig. 8 summarizes the results of this analysis, that alsoshows the line-drawings of the final step of all the experi-ments (fig. 8, row 1). Notice that the drawings arefocused on the central part of the sandboxes, unaffectedby edge effects. A related sequence of diagrams is shown

below each of the line-drawings. All values reported in thediagrams of rows 2 and 3 were measured on picturestaken at different deformation stages (every 0.5 cm ofbasal displacement). Finally, values reported in the dia-grams of row 4 were measured at the surface of the ana-logue models at the end of the experiments. All displace-ment measurements have been based on the «referencelines» traced on the surface of all the models (named bycapital letters; fig. 8, row 1).

5.1. MAXIMUM DISPLACEMENT ALONG THE FAULTS (fig. 8,row 2).

These diagrams show the displacement on each faultsurface measured at different steps of deformation. Themeasurements were taken along the fault trace and moni-tor the increase of displacement every 0.5 cm of baseplatefault slip. The total displacement imposed on the base-plate fault is indicated on the x-axis, whereas the corre-sponding displacement measured on every single fault isreported on the y-axis. These diagrams allow tracing theentire evolution of the faults in the models: inception,progressive development, deactivation.

5.2. EVOLUTION OF THE SURFACE DEFORMATION (fig. 8,row 3).

These diagrams show the cumulative horizontal dis-placement of all faults (y-axis) measured at each refer-ence line (x-axis) for each step of deformation (differentlines). The measurements were taken parallel to the direc-tion of the baseplate fault, regardless of the strike of thefaults displacing the reference lines. These diagramsallow quantification of how much of the imposed basaldisplacement is transferred to the model surface wherethe pre-existing cut affects the entire thickness of themodels (right side of each diagram) or where it affectsonly part of them, tapering to zero below the orogenicwedge (left side of each diagram).

5.3. CONTRIBUTION OF EACH FAULT TO THE TOTAL DIS-PLACEMENT (fig. 8, row 4).

These diagrams refer to the final step of the models(due to the extremely limited number of measurementswe did not diagram model SS06) and show the number of


Fig. 7 - Sketched 3D interpretation of the central partof the model SS05 (D=3.0 cm). The curved planes ofnewly-formed faults are shown in light grey, whereasthe cut in the foreland both exposed and buried belowthe wedge is shown in dark grey. – Interpretazione 3D schematica della parte centraledel modello SS05 (D=3.0 cm). I piani curvi delle faglieneo-formate sono rappresentati in grigio chiaro, mentreil taglio nell’avampaese, sia affiorante sia sepolto sottoil cuneo, è mostrato in grigio scuro.

faults and the contribution of each fault to the total dis-placement, measured using the reference lines. The verti-cal bars reproduce the path of the uppermost curve in thediagrams of row 3. Numbers inside the bars refer to thefault numbers in the maps of the models shown in row 1.

6. DISCUSSION

6.1. GENERAL FEATURES IN 3D

An overview of the three domains shows that newly-formed faults occur only in domains B and C, as only theinherited discontinuity exists and is active in domain A.Whatever their geometry and displacement, all thesefaults splay from a pre-existing lineament (the cut indomain B, the baseplate fault in domain C). The three-dimensional sketch of model SS05, shown in figs. 6 and7, provides an integrated view of what is seen both at thesurface and in the cross sections. Significant features are:(i) the general arrangement of the faults, mainly in thenorthern block; (ii) the lack of new faults where the pre-existing fault involves the entire thickness of the model,i.e. in domain A; (iii) the presence of faults only withinthe wedge where it overlies the foreland, i.e. in domain B;and (iv) faults splaying directly from the baseplate faultwhere no other lineaments or heterogeneities exist, that isto say, in domain C. The main faults cutting the model (inlight grey in fig. 7) are connected with the baseplate faultin domain C and with the cut in domain B. Toward thesurface, they depart from the vertical projection of thesepre-existing discontinuities; the undulate shape of thefault planes is outlined by their arched traces, both inmap view and in cross section. Toward domain B, slip onthe pre-existing cut propagated only in the recedingblock, producing distortion of the grid lines, precursor tothe development of faults. This behavior is easy to under-stand qualitatively, since the material is compressed onthe advancing side of the fault and stretched on the reced-ing side (MANDL, 2000). Conversely, grid lines remainalmost undeformed to the south of fault P (figs. 5 and 6).

From east to west, the influence of the pre-existingcut diminishes as the thickness of the overburden (i.e., ofthe chain) increases. Accordingly, the width of the shearzone measured at the surface increases toward the west.In domain C, that has no pre-existing discontinuity, thebaseplate fault controls completely the structural style ofthe overburden, and the shear zone attains its maximumwidth where the overlying sand pack is thickest. Similarobservations were reported also by TCHALENKO (1970)and SCHÖPFER & STEYRER (2001). Moreover, grid linesbegin to be significantly deformed long before faultsreach the surface.

6.2 QUANTITATIVE ANALYSIS

In the following we discuss the finer details of thequantitative analysis, always referring to the differentrows of fig. 8.

6.2.1. Maximum displacement along faults (fig. 8, row 2).

In all models, the number of simultaneously activefaults becomes larger as the amount of basal displace-ment increases. Only two faults are active in addition tofault P (the pre-existing cut in domain A) when the

imposed displacement is less than 1.0-1.5 cm. Recall thatfault P is always active in all the models and shows dis-placement values very close to those of the baseplatefault. This indicates that deformation in domain A isentirely transferred to the model surface only by thisfault. Conversely, an increase in the number of activefaults is seen in all the models for basal displacement>1.5 cm. These faults remain active for a significant inter-val of time/basal-displacement: up to ~6 cm of basal dis-placement in model SS03, and up to the end of the exper-iments in models SS04 and SS05.

In models SS03 and SS04, at ~5.5 cm of basal dis-placement the surface faults striking close to the base-plate fault join one another forming a single larger fault.This is directly linked with the baseplate fault, extends forthe total length of the models and divides them into twoseparate halves. The central lineament is the only onethat remains active for displacements larger than 5.5 cm.Hence the threshold of basal displacement beyond whichany increase in displacement occurs essentially on themain fault can be set between 5.0 and 6.0 cm, whereaschanges in the surface structures are negligible.

6.2.2. Surface deformation evolution (fig. 8, row 3).

All diagrams show similar paths. The reference linesof domain A (from O to S) exhibit displacement valuesthat are very close to the corresponding baseplate faultdisplacement. This is due to the reactivation of the pre-existing discontinuity, which accommodates the entiredeformation thus preventing the development of newfaults. Conversely, in domain B and, even more, indomain C, the measured horizontal surface displace-ments differ from those of the corresponding baseplatefault; in particular, they decrease abruptly starting fromthe reference line at the boundary between domains Aand B (fig. 8, row 3, reference line O). This difference canbe explained only by the occurrence of distributed defor-mation (layer-parallel shortening, large-wavelength fold-ing, fault propagation folding, brittle deformation notresolvable at the model scale, etc.) and of a minor reversecomponent of motion on mainly strike-slip faults. Thepartitioning of deformation cannot be quantified basedon the analysis of the surface reference lines, but can beinferred from the analysis of the sections perpendicularto the baseplate fault, that show positive variations in thetopography of the models. Indeed, topographic variationshave been detected in several natural and experimentalenvironments (SYLVESTER, 1988; LE GUERROUÉ & COB-BOLD, 2006). In particular, the development of Riedelfault systems or the formation of flower structures areusually associated with bulging confined within narrowportions of the models or of real shear zones, the SanAndreas Fault in the Carrizo Plain being the most strikingand widely known example (U.S.G.S., 1990). In the mod-els analyzed in this paper, such topographic bulge charac-terizes mainly domain C, as shown by the sections in fig.6 (domain C, S 34 and S 26). The examination of theentire set of available sections (not shown in figure) sug-gests that bulging of domain C corresponds to the maxi-mum width of the shear zones for all models.

The portion of basal displacement that is partitionedas described above, and ultimately not transferred to themodels surface, is quantified by the diagrams of fig. 8,row 3. No brittle deformation is seen at the surface of



A

Fig. 8a - Diagrams referring to models SS03 (D=8.0 cm, first column) and SS04 (D=5.0 cm, second column). Row 1: line-drawing of the final stage of the models. Row 2: maximum displacement along the faults. Row 3: evolution of surfacedeformation. Row 4: contribution of each fault to the total displacement. Numbers and letters in the diagrams of row 4refer to the line-drawings of row 1. See text for details about measurements.– Diagrammi relativi ai modelli SS03 (D=8.0 cm, prima colonna) e SS04 (D=5.0 cm, seconda colonna). Riga 1: line-drawingdella fase finale del modello. Riga 2: massimo rigetto lungo le faglie. Riga 3: evoluzione della deformazione superficiale. Riga 4:contributo delle singole faglie al rigetto totale. I numeri e le lettere nei diagrammi nella riga 4 si riferiscono al line-drawingdella riga 1. Per maggiori dettagli si veda il testo.


B

Fig. 8b - Diagrams referring to models SS05 (D=3.0 cm, first column) and SS06 (D=0.5 cm; second column). Row 1: line-drawing of the final stage of the models. Row 2: maximum displacement along the faults. Row 3: evolution of surfacedeformation. Row 4: contribution of each fault to the total displacement. Numbers and letters in the diagrams of row 4refer to the line-drawings of row 1. See text for details about measurements.– Diagrammi realtivi ai modelli SS05 (D=3.0 cm, prima colonna) e SS06 (D=0.5 cm, seconda colonna). Riga 1: line-drawingdella fase finale del modello. Riga 2: massimo rigetto lungo le faglie. Riga 3: evoluzione della deformazione superficiale. Riga 4:contributo delle singole faglie al rigetto totale. I numeri e le lettere nei diagrammi nella riga 4 si riferiscono al line-drawingdella riga 1. Per maggiori dettagli si veda il testo.

domain C (reference lines from C to G) for basal displace-ment less than ~2.0 cm. On the contrary, surface faultsare clearly seen in domain B at the same deformationstage. As stated earlier, in domain A the surface and basaldisplacements are very similar. Conversely, during theearly deformation stages (basal displacement <4.0 cm) thesurface displacement accommodated by faults in domainsB and C is consistently smaller than the correspondingsteps of basal displacement. For instance, in model SS03(fig. 8, row 3, first diagram) the lines corresponding to 3.0and 4.0 cm of basal displacement are clearly less than 1.0cm apart in domain C (reference lines E, F and G), whilethey are spaced exactly 1.0 cm in domain A.

For displacements larger than 6.0 cm, the differencebetween the lines is consistently the same (1.0 cm) overall diagrams and for all domains. This is due the fact thatfor any displacement >5.5 cm all faults join and the

model deformation reaches a sort of «steady-state». Sincethat moment, a single active main fault separates com-pletely the overburden up to the surface into two blockscorresponding with the two baseplate halves. Conversely,for displacement <5.5 cm diffuse deformation still occursin domains B and C and accommodates a fraction of thebaseplate fault displacement roughly corresponding tothe difference between the imposed and measured dis-placements. Model SS03, that was run with a basal dis-placement of 8.0 cm, is particularly suitable for thesemeasurements as it clearly reached such steady-state. Themaximum difference between imposed and measured dis-placement is 2.5 cm and can be observed at lines G-H,that separate domains B and C (fig. 8, row 3, first dia-gram). A difference between imposed and measured dis-placements can be observed also for the other models,when basal displacement grows beyond 2.5 cm (SS04 and


Fig. 9 - Surface evolution of the reference model SS02 and measured amount of displacement on the different reference lines taken at six dif-ferent steps of deformation (from 3 to 8 cm of basal displacement).– Evoluzione superficiale del modello di riferimento SS02 ed entità dei rigetti misurata sulle linee di riferimento a sei differenti step di deformazione(da 3 a 8 cm di rigetto basale).

SS05; fig. 8, row 3, second and third diagrams) whichcorresponds to the minimum displacement needed toallow faulting to affect the surface of all domains.

Moving to the westernmost portion of the models(domain C), the measured displacement increases again,for instance at reference lines E, D and C. This increase isconcomitant with the progressive decrease of the width ofthe shear zone (fig. 8, row 1). Moreover, sections of thesame portion of the models show a progressive reductionof the topographic relief, indicating that part of the dis-tributed or non-detectable deformation is also decreasingand the difference between basal and surface displace-ment diminishes.

In model SS03, near the boundary between domains Band C the reduction of measured surface displacement ismore significant (2.5 cm). As the imposed displacement isconstant, deformation in this portion of the model does notconcentrate along fault planes but appears rather spreadout. Displacement increases again as one moves westwardfrom the wedge boundary. Therefore, the concomitantpresence of the overlying wedge and the pre-existing faultin the buried foreland seem to play a role in reducing thedevelopment of faults and partitioning the deformation inthe frontal part of domain B. A detailed analysis of thisdeformation style could be the object of future studies.

The same measurements have been carried out forthe reference model SS02 (fig. 9). The lack of discontinu-ities is the essential reason for the obvious differences inthe fault pattern. These differences are also highlightedby the measurements in the diagram at the bottom of thefigure, showing that the distribution of faults and thestrain partitioning are quite constant for the whole dura-tion of deformation and all along the model.

6.2.3. Contribution of each fault to the total displacement.

The diagrams of fig. 8, row 4, illustrate the predomi-nant contribution of the pre-existing fault in accommodat-ing the displacement in domain A and in the eastern partof domain B (near the orogenic front, where the wedge isthinner). Conversely, a larger number of faults is seen alsofor a smaller cumulated displacement away from the fore-land, where the wedge is thicker, and in the chain domainC. In all models, faults take up a larger displacement whentheir strike is more similar to that of the baseplate fault(e.g. fault 5a in SS03 model; fig. 8, row 4, first diagram)and viceversa (e.g. faults 1 and 2 in models SS03 andSS04; fig. 8, row 4, first and second diagrams). Notice thatfaults stemming directly from the pre-existing cut increasetheir displacement abruptly from zero to large values,probably as a result of displacement being transferredfrom the main onto these newly-formed faults. Notice alsothat the maximum displacement accommodated by eachnewly-formed fault does not exceed ~3.5 cm, which indeedgives an indication of how efficient is the reactivation ofthe pre-existing discontinuity.

Based on the observations and quantitative estimatesdiscussed above, we can now re-consider the sandboxmodels in terms of modes of crack propagation (fig. 1).The boundary between domains A and B (in correspon-dence with the pre-existing strike-slip fault) is the tip dam-age zone (sensu KIM et alii, 2006), so that Mode-II isprevalent. As previously stated, the effects of the inheritedlineament are almost unobserved in domain C. Mode-III isdominant in this part of the models, as suggested also by

the symmetrical arrangement of the newly-formed faults.The shift from one mode to the other approximatelyoccurs in domain B (the wedge) where the effects of thepre-existing lineament decrease going toward domain C.

7. CONCLUSIONS

Qualitative and quantitative analyses performed onsandbox models allowed us to verify that the reactivationof a pre-existing lineament strongly controls the develop-ment and arrangement of newly-formed faults in anundeformed cover, in our case the frontal part of an oro-genic wedge overlying its buried foreland. In particularwe found that:

– in the foreland domain A, the pre-existing fault actsas a preferential slip surface accommodating almost allthe displacement so that no new faults are formed. In fig.8, row 2, the line representing fault P (pre-existing) is thediagonal of the chart, as the basal and surface displace-ments are equal. The same is seen in fig. 8, row 4; the barheight corresponds to the basal displacement in the fore-land zone (domain A), and the bars refer always to a sin-gle fault (P).

– in the wedge domain B, near the front of the oro-genic wedge, faults are asymmetrically distributed andexclusively located in the receding block; the effect of thereactivated pre-existing lineament on surface displace-ment decreases moving away from the foreland-wedgeboundary. Fig. 8, row 3 shows that the surface displace-ment measured in this domain (reference lines N, M, L, I)is the smallest of the entire model. This feature is seen foreach step of model development and for all models.

– in the chain domain C, faults are widespread andexhibit features that are similar to those widely describedin literature. As shown by the evolution diagrams of thefour analyzed models (fig. 8, row 1), faults cutting thisdomain are the last to develop.

In summary, newly-formed faults take place only intwo domains:

– in domain B, where all the newly-formed faultsoriginate from the pre-existing vertical fault that exists inthe foreland buried under the wedge;

– in domain C, where the faults splay up from thebaseplate fault.

New faults forms only within the undeformed cover.Conversely, no new faults take place where the pre-exist-ing lineament involves the entire thickness of the modelup to the surface (domain A). In this case, the pre-existingfault accommodates the entire basal displacement sincethe very early stages of deformation.

Surface faulting shows different characteristics in thethree domains. Domain A is the first one where progres-sive displacement appears at the surface; then, for increas-ing basal displacement domains B and C are also affectedby faults. Slip at the surface always corresponds to thebasal displacement in domain A, whereas in the other twodomains the horizontal slip measured at surface slip isonly part of the corresponding basal displacement. This isdue to the occurrence of diffuse deformation and minorreverse components of motion on strike-slip faults. For dis-placements >5.5 cm, all faults in the central part of themodels join and deformation reaches a «steady-state».


From this stage onwards, a single active main fault sepa-rates the entire overburden into two separate blocks.

The three domains also differ for the number offaults: only the pre-existing fault in domain A, few faultsin domain B, more faults in domain C.

Finally, faults having strike comparable to the base-plate fault take up a larger displacement and remainactive for a longer time, whereas more oblique faults withremain active for a shorter time and their displacementmeasured along strike of the main fault is smaller.

APPENDIX

Comparison with geological cases

In this appendix, we compare our results with two nat-ural examples in Italy: the Molise-Gondola (hereinafterMGsz) and Scicli-Ragusa shear zones (figs. 10 and 11).The first case study is a reminder about an E-W strikingdeformation belt that cuts the Adriatic foreland (DI BUCCI

et alii, 2006; 2007; fig. 10), and it is aimed at comparingthe structural setting of the models with the seismotec-tonic setting of the considered area. The central-easternportion of the MGsz is exposed at the surface, whereas itswestern portion is buried below the Apennine foredeepand the outer front of the Southern Apennines. Evidencefor MGsz activity dates back to the Cretaceous, but thisstructure is still active as a right-lateral shear zone, asdemonstrated by the frequent and sometimes destructiveassociated seismicity. Notice that the Southern Apenninesthrust sheets and the related foredeep deposits buried thewesternmost part of the MGsz in Middle Pleistocene,

when motion at the front of the chain ceased (PATACCA &SCANDONE, 2004b). Any subsequent motion along theMGsz must have propagated through the overlying coveras described earlier. Our second case study is from South-eastern Sicily (fig. 11). Here the Hyblean foreland, affectedby a long-lived shear zone known as Scicli-Ragusa faultsystem, plunges beneath the foredeep and the front of theGela nappe thrust sheets. Also this case fits with the gen-eral framework of our models. Both our case studies com-prise very active areas that are not fully understood fromthe seismotectonic point of view.

The quantification and the minute description of thestructural features observed in the models and the resultspresented in this paper can be used both to interpret dif-ferent geological settings, and to shed light on differentgeological settings and put forward hypotheses on the tec-tonic setting of areas where field studies are still at anearly stage and need to be properly oriented. In particu-lar, the integration of the results of this work with exist-ing field geology studies allowed us:

(i) to provide an interpretative key for specific char-acteristics of the MGsz;

(ii) to attempt to use our models for the seismotec-tonic interpretation of a comparable pre-existing forelandshear zone in Southeastern Sicily.

INTERPRETING SPECIFIC CHARACTERISTICS OF THE MGSZ

The MGsz has been described as a system that can beidentified over a distance of at least 180 km running atthe latitude 41°40’N approximately. Overall the MGszappears as a ~15 km-wide corridor from the Adriatic fore-


Fig. 10 - Simplified sketch of the Central-Southern Apennines chain (light grey) and related foreland (dark grey), centered on the Mattinata-Gondola shear zone (MGsz). Focal mechanisms of earthquakes with M > 5 that occurred along the MGsz are also shown (after GASPARINI etalii, 1985, DI LUCCIO et alii, 2005).– Schema semplificato della catena Appenninica Centro-Meridionale (grigio chiaro) e del relativo avampaese (grigio scuro), focalizzata sulla Mat-tinata-Gondola shear zone (MGsz). Vengono mostrati i meccanismi focali dei terremoti con M>5 lungo la MGsz (da GASPARINI et alii, 1985, DILUCCIO et alii, 2005).

land off-shore to the core of the Apennines fold-and-thrust belt (fig. 10). Its off-shore portion is known as Gon-dola line (DE’ DOMINICIS & MAZZOLDI, 1987; COLANTONI

et alii, 1990; DE ALTERIIS, 1995; MORELLI, 2002; RIDENTE

et alii, 2008; fig. 10), whereas its onshore portion includesthe Mattinata fault (FINETTI, 1982; ORTOLANI &PAGLIUCA, 1987; FUNICIELLO et alii, 1988; PICCARDI, 1998;WINTER & TAPPONIER, 1991; BILLI & SALVINI, 2000;CHILOVI et alii, 2000; BILLI, 2003; BORRE et alii, 2003;PICCARDI, 2005; TONDI et alii, 2005; BILLI et alii, 2007)and the source regions of the 30 July 1627 Gargano earth-

quake sequence (Mw 6.7: Gruppo di Lavoro CPTI; 2004;PATACCA & SCANDONE, 2004a) and of the 31 October-1November 2002 Molise earthquakes (both having Mw 5.7:DI LUCCIO et alii, 2005; VALLÉE & DI LUCCIO, 2005; fig.10). Our analogue models confirm that the eastern por-tion of the MGsz, coincident with the Gondola and Matti-nata fault systems, has a good chance to be fully reacti-vated up to the surface (as proposed by PICCARDI, 1998;BORRE et alii, 2003; PICCARDI, 2005; TONDI et alii, 2005;RIDENTE & TRINCARDI, 2006, among others) also forextremely small displacement values.


Fig. 11 - a) Schematic structural setting of southeastern Sicily. Data simplified and redrawn after LENTINI ed. (1984), BIANCHI et alii, (1987),BIGI et alii, (1992), D’AGOSTINO & SELVAGGI (2004). b) The three main structural domains have been colored as in the sandbox models. – a) Schema strutturale semplificato della Sicilia Sud-orientale. Dati presi e rielaborati da LENTINI ed. (1984), BIANCHI et alii (1987), BIGI et alii(1992), D’AGOSTINO & SELVAGGI (2004). b) I tre principali domini strutturali sono stati colorati coerentemente con quanto fatto per i modellianalogici.

To the west of the Mattinata fault, the Apricena fault(the causative fault of the 1627 earthquakes according toPATACCA & SCANDONE, 2004a) has been interpreted by DI

BUCCI et alii (2006) as one of the splay faults developingat the front of the orogenic wedge from the deeper, pre-existing discontinuity in domain B of our models. Recallthat these splays strike N288° and dip to the SSW, exhibita normal component of slip and form even for relativelylow displacements.

The sources of the 31 October-1 November 2002Molise earthquakes are steeply dipping, blind right-lateralstrike-slip faults having a cumulative length of about 15km and extending between 6 and 20 km depth (VALLÉE &DI LUCCIO, 2005; DISS Working Group, 2007). Despitethe limited size of the events, GPS data revealed coseis-mic deformation consistent with dextral kinematics (GIU-LIANI et alii, 2007). The drainage pattern of the regionoverlying the 2002 seismogenic faults displays anomalousriver sections; some of them trend E-W, violating the gen-eral SW-NE trend of drainage (VALENSISE et alii, 2004).Accordingly, our experiments show that in the portion ofdomain B that corresponds to the structural setting of the2002 earthquakes, the pre-existing strike-slip fault doesnot reach the model surface until displacement exceeds~5.5 cm, corresponding to ~10 km in nature. Neverthe-less, the model surface is affected by a region of diffusestrain, showing both horizontal and vertical componentsof motion, even for very limited displacements.

In a more general perspective, the comparisonbetween our models and the structural setting of theMGsz further highlights that the complex structural pat-tern generated by large fault displacements does not havean obvious equivalent in the modeled part of the Apen-nines. This observation favors the hypothesis that themost recent activity of the MGsz has not yet accumulateda significant displacement.

Finally, we can use our results and reasoning to spec-ulate on the style of potentially active faults west of thesource area of the 2002 Molise earthquakes.

To do that, we move further toward the interior of theApennines, i.e. toward the western edge of the models.Here the right-lateral, deep vertical fault is accompaniedby moderately- to steeply-dipping oblique faults that mayor may not reach the surface depending on the horizontaloffset; for instance, no surface faulting is predicted fordisplacement values such as those discussed above (~1km in nature). The structural setting of potential activestructures west of the 2002 Molise earthquakes shouldhence be characterized by deep faulting and diffuse sur-face deformation, without primary surface faulting. Thisconclusion agrees with the reported lack of surface activefaults in this area (CIARANFI et alii, 1983; GALADINI et alii,2000; MICHETTI et alii, 2000; DISS Working Group, 2007).

INTERPRETING THE SCICLI-RAGUSA SHEAR ZONE (SOUTH-EASTERN SICILY)

The satisfactory match between the outcomes of ourmodeling and the setting of the MGsz suggests that theresults of our work could be applied to other long-livedstrike-slip zones having a geological setting comparableto the one that inspired our models but not fully under-stood from the neotectonic point of view. This is the caseof the southeastern Sicily (fig. 11), where a portion of theforeland exposed in the Hyblean Mts. plunges northward

beneath the Gela foredeep and the external front of theGela nappe. This front runs from the Gela foredeep to theCatania Plain and Mt. Etna (LENTINI ed., 1984; BIANCHI

et alii, 1987; CARBONE et alii, 1987; BIGI et alii, 1992; fig.11). The Hyblean Mts. are cut by a regional shear zoneknown as Scicli-Ragusa fault system, a large structuralfeature that developed in Meso-Cenozoic times with apredominant right-lateral component of motion (e.g.,GHISETTI & VEZZANI, 1980; CARBONE et alii, 1987, withreferences). The current stress field, however, is charac-terized by NNW-SSE contraction related to the Eurasia-Nubia convergence (D’AGOSTINO & SELVAGGI, 2004;MONTONE et alii, 2004; SERPELLONI et alii, 2007; fig. 11).Evidence for NNW-oriented contraction is supplied alsoby recent structural observations (CATALANO et alii, 2006;LAVECCHIA et alii, 2007); hence, similarly to the MGsz, theScicli-Ragusa fault system is favorably oriented for reacti-vation with an opposite sense of motion, i.e. as a left-lat-eral shear zone (CATALANO et alii, 2008), a circumstancesupported also by focal mechanism data (MUSUMECI etalii, 2005).

Our models suggest that evidence of recent motionfor this fault system should be found not only where it isexposed, but also where it is buried under the Gela fore-deep and the Gela nappe. Here we should see deeper seis-micity concentrated along the northern extension of theScicli-Ragusa fault system, whereas near or at the surfacewe might expect transtensional faults turning anticlock-wise from the surface trace of the line. Current catalogues(Gruppo di Lavoro CPTI, 2004) show a concentration ofhistorical earthquakes along the northward prolongationof the Scicli-Ragusa fault system, but accurate surfacestructural data (field surveys, mesostructural analyses,absolute dates; a preliminary set of fresh observations issupplied by CATALANO et alii, 2006; 2008) are still stronglyneeded. New data would highlight the presence or theabsence of structural features like those described in thispaper, hopefully supporting our work hypotheses andpossibly supplying valuable insight into the seismotecton-ics of Sicily and the dynamics of Mt. Etna.

Acknowledgments

This work was supported by the University of Pavia, the IstitutoNazionale di Geofisica e Vulcanologia and the Italian Dipartimentodella Protezione Civile. The authors are grateful to Luca Viscontiand Claudio Antoniazzi, who provided part of the measurements onthe models. Quantitative analysis were performed using the software«MOVE» by Midland Valley Exploration Ltd., Glasgow, UK. TheEditor and the referees Fernando Calamita and Giacomo Corti arekindly acknowledged for the thorough review and the constructivecriticisms that greatly improved this paper.


TABLE 1

– List of the experiments described in this study and oftheir geometrical parameters.

– Lista degli esperimenti descritti in questo studio e dei loro parametri geometrici.

Pre-existingLayer of glass

PresenceExperimentcut

microbeads of the wedge

Thickness Displacementin the foreland

SS02 No No No 10 cm 8.0 cmSS03 Yes Yes Yes 10-11 cm 8.0 cmSS04 Yes Yes Yes 10-11 cm 5.5 cmSS05 Yes Yes Yes 10-11 cm 3.0 cmSS06 Yes Yes Yes 10-11 cm 0.5 cm

REFERENCES

BIANCHI F., CARBONE S., GRASSO M., INVERNIZZI G., LENTINI F.,LONGARETTI G., MERLINI S. & MOSTARDINI M. (1987) - SiciliaOrientale: profilo Nebrodi-Iblei. Memorie della Società GeologicaItaliana, 38, 429-458.

BIGI G., COSENTINO D., PAROTTO M., SARTORI R., SCANDONE P.(1992) - Structural Model of Italy and Gravity Map. CNR, P.F.Geodinamica, Quaderni de «La Ricerca Scientifica» 114 (3).

BILLI A. (2003) - Solution slip and separations on strike-slip fault zo-nes: theory and application to the Mattinata Fault, Italy. Journalof Structural Geology, 25, 703-715.

BILLI A. & SALVINI F. (2000) - Sistemi di fratture associati a faglie inrocce carbonatiche: nuovi dati sull’evoluzione tettonica del Pro-montorio del Gargano. Boll. Soc. Geol. It., 119, 237-250.

BILLI A., GAMBINI R., NICOLAI C. & STORTI F. (2007) - Neogene-Qua-ternary intraforeland transpression along a Mesozoic platform-ba-sin margin. The Gargano fault system, Adria, Italy. Geosphere, 3(1), 1-15.

BORRE K., CACON S., CELLO G., KONTNY B., KOSTAK B., LIKKE AN-DERSEN H., MORATTI G., PICCARDI L., STEMBERK J., TONDI E. &VILIMEK V. (2003) - The COST project in Italy: analysis and moni-toring of seismogenic faults in the Gargano and Norcia areas (cen-tral southern Apennines, Italy). Journal of Geodynamics, 36, 3-18.

BYERLEE J. (1978) - Friction of rocks. Pure and applied geophysics,116, 615-626.

CARBONE S., GRASSO M. & LENTINI F. (1987) - Lineamenti geologicidel Plateau Ibleo (Sicilia SE). Presentazione delle carte geologichedella Sicilia sud-orientale. Memorie della Società Geologica Ita-liana, 38, 127-135.

CASAGRANDE A. (1932) - Research on the Atterberg limits of soils. Pu-blic Roads, 13 (9), 121-136.

CATALANO S., DE GUIDI G., LANZAFAME G., MONACO C., TORRISI S.,TORTORICI G. & TORTORICI L. (2006) - Inversione tettonica posi-tiva tardo-quaternaria nel Plateau Ibleo (Sicilia SE). Rendicontidella Società Geologica Italiana, 2 (2006), 118-120.

CATALANO S., DE GUIDI G., ROMAGNOLI G., TORRISI S., TORTORICI G.& TORTORICI L. (2008) - The migration of plate boundaries in SESicily: Influence on the large-scale kinematic model of the Africanpromontory in southern Italy. Tectonophysics, 449 (1-4), 41-62.

CHILOVI C., DE FEYTER A.J. & POMPUCCI A. (2000) - Wrench zonereactivation in the Adriatic Block: the example of the MattinataFault System (SE Italy). Bollettino della Società Geologica Italia-na, 119 (1), 3-8.

CHRISTIE-BLICK N. & BIDDLE K.T. (1985) - Deformation and basinformation along strike-slip faults. In: Biddle K.T. & Christie-Blick N., eds., Strike-slip deformation, basin formation, and se-dimentation, Society of Economic Paleontologists and Minera-logists, Special Publication, 37, 1-34.

CIARANFI N., GHISETTI F., GUIDA M., IACCARINO G., LAMBIASE S.,PIERI P., RAPISARDI L., RICCHETTI G., TORRE M., TORTORICI L.& VEZZANI L. (1983) - Carta neotettonica dell'Italia meridionale.C.N.R., Publ. 515 of Progetto Finalizzato Geodinamica.

COLANTONI P., TRAMONTANA M. & TEDESCHI R. (1990) - Contributoalla conoscenza dell’avampaese apulo: struttura del Golfo di Man-fredonia (Adriatico meridionale). Giornale di Geologia, serie 3a,52 (1-2), 19-32.

D’AGOSTINO N. & SELVAGGI G. (2004) - Crustal motion along the Eu-rasia-Nubia plate boundary in the Calabrian Arc and Sicily andactive extension in the Messina Straits from GPS measurements.J. Geophys. Res., 109, B11402, doi:10.1029/2004JB002998.

DE ALTERIIS G. (1995) - Different foreland basins in Italy: examplesfrom the central and southern Adriatic Sea. Tectonophysics, 252,349-373.

DE’ DOMINICIS A. & MAZZOLDI G. (1987) - Interpretazione geologico-strutturale del margine orientale della Piattaforma apula. Memo-rie Società Geologica Italiana, 38, 163-176.

DI BUCCI D., RAVAGLIA A., SENO S., TOSCANI G., FRACASSI U. & VALENSISE G. (2006) - Seismotectonics of the Southern Apenni-nes and Adriatic foreland: insights on active regional E-W shearzones from analogue modeling. Tectonics, 25, TC4014, doi:10.1029/2005TC001898.

DI BUCCI D., RAVAGLIA A., SENO S., TOSCANI G., FRACASSI U. & VA-LENSISE G. (2007) - Modes of fault reactivation from analoguemodeling experiments: implications for the seismotectonics of thesouthern Adriatic foreland (Italy). Quaternary International, 171-172, 2-13, doi: 10.1016/j.quaint.2007.01.005.

DI LUCCIO F., FUKUYAMA E. & PINO N.A. (2005) - The 2002 Moliseearthquake sequence: What can we learn about the tectonics ofsouthern Italy?. Tectonophysics, 405, 141-154.

DISS WORKING GROUP (2007) - Database of Individual SeismogenicSources (DISS), Version 3.0.3: A compilation of potential sour-ces for earthquakes larger than M 5.5 in Italy and surroundingareas. http://www.ingv.it/DISS/, © INGV 2005, 2007. All rightsreserved.

FINETTI I. (1982) - Structure, stratigraphy and evolution of the CentralMediterranean. Boll. Geofis. Teor. Appl., 24, 247-312.

FUNICIELLO R., MONTONE P., SALVINI F. & TOZZI M. (1988) - Carat-teri strutturali del Promontorio del Gargano. Mem. Soc. Geol. It.,41, 1235-1243.

GASPARINI C., IANNACCONE G. & SCARPA R. (1985) - Fault-plane solu-tions and seismicity of the Italian peninsula. Tectonophysics,117, 59-78.

GALADINI F., MELETTI C. & VITTORI E. (2000) - Stato delle conoscen-ze sulle faglie attive in Italia: elementi geologici di superficie. In:Galadini et alii (eds), «Le ricerche del GNDT nel campo della pe-ricolosità sismica (1996-1999)», CNR-Gruppo Nazionale per laDifesa dai Terremoti - Roma, 107-136 (also available from:http://www.ingv.it/gndt/Pubblicazioni/Meletti_copertina.htm).

GHISETTI F. & MEZZANI L. (1980) - The structural features of theIblean plateau and of the mount Judica area (south-eastern Si-cily): a microtectonic contribution to the deformational history ofthe Calabrian Arc. Boll. Soc. Geol. It., 99, 57-102.

GIULIANI R., ANZIDEI M., BONCI L., CALCATERRA S., D’AGOSTINO N.,MATTONE M., PIETRANTONIO G., RIGUZZI F. & SELVAGGI G.(2007) - Co-seismic displacements associated to the Molise(Southern Italy) earthquake sequence of October-November 2002inferred from GPS measurements. Tectonophysics, 432, 21-35.

GRUPPO DI LAVORO CPTI (2004) - Catalogo Parametrico dei TerremotiItaliani, versione 2004 (CPTI04). INGV, Bologna, http://emi-dius.mi.ingv.it/CPTI/

HARDING T.P. (1985) - Seismic characteristics and identification of ne-gative flower structures, positive flower structures, and positivestructural inversion. The American Association of PetroleumGeologists Bulletin, 69 (4), 582-600.

HUBBERT M.K. (1937) - Theory of scale models as applied to the studyof geological structures. Geological Society of America Bulletin,48, 1459-1520.

KIM Y.-S., PEACOCK D.C.P. & SANDERSON D.J. (2004) - Fault damagezones. Journal of Structural Geology, 26, 503-517.

KIM Y.-S. & SANDERSON D.J. (2006) - Structural similarity and va-riety at the tips in a wide range of strike-slip faults: a review. TerraNova, 18, 330-344.

LAVECCHIA G., FERRARINI F., DE NARDIS R., VISINI F. & BARBANO

M.S. (2007) - Active thrusting as a possible seismogennic sourcein Sicily (Southern Italy): some insights from integrated structu-ral-kinematic and seismological data. Tectonophysics, 445, 145-167.

LE GUERROUÉ E. & COBBOLD P.R. (2006) - Influence of erosion andsedimentation on strike-slip fault systems: insights from analoguemodels. Journal of Structural Geology, 28, 421-430.

LENTINI F. (Ed.) (1984) - Carta geologica della Sicilia Sud-Orientale,scale 1:100.000. Published by Selca, Firenze.

MANDL G. (1988) - Mechanics of tectonic faulting: models and basicconcepts. Netherlands, Elsevier, 407 pp.

MANDL G. (2000) - Faulting in Brittle Rocks. 434 pp., Springer, Ber-lin.

MICHETTI A.M., SERVA L. & VITTORI E. (2000) - ITHACA. Italy Ha-zard from Capable Faults: a database of active faults of the Italianonshore territory. Cd-rom and explication notes, published byANPA, Rome.

MONTONE P., MARIUCCI M.T., PONDRELLI S. & AMATO A. (2004) - Animproved stress map for Italy and surrounding regions (centralMediterranean). Journal of Geophysical Research, 109(B10410), 1-22, doi:10.1029/1003JB002703.


MORELLI D. (2002) - Evoluzione tettonico-stratigrafica del MargineAdriatico compreso tra il Promontorio garganico e Brindisi. Me-morie Società Geologica Italiana, 57, 343-353.

MUSUMECI C., PATANE D., SCARFI L. & GRESTA S. (2005) - Stress di-rections and shear-wave anisotropy: observations from localearthquakes in Southeastern Sicily, Italy. Bulletin of the Seismo-logical Society of America, 95 (4), 1359–1374.

NAYLOR M.A., MANDL G. & SIJPESTEIJN C.H.K. (1986) - Fault geome-tries in basement-induced wrench faulting under different initialstress states. Journal of Structural Geology, 8, 737-752.

NIEUWLAND D.A. & NIJMAN M. (2001) - The atlas of structural geome-try: a digital collection of 25 years of analogue modelling. Geolo-gie en Mijnbouw/Netherlands Journal of Geosciences, 80, 59-60.

ORTOLANI F. & PAGLIUCA S. (1987) - Tettonica transpressiva nel Gar-gano e rapporti con le Catene Appenninica e Dinarica. Mem. Soc.Geol. It., 38, 205-224.

PATACCA E. & SCANDONE P. (2004a) - The 1627 Gargano earthquake(Southern Italy): Identification and characterization of the causa-tive fault. Journal of Seismology, 8 (2), 259-273.

PATACCA E. & SCANDONE P. (2004b) - The Plio-Pleistocene thrust belt– foredeep system in the Southern Apennines and Sicily (Italy). In:Crescenti U., D’Offizi S., Merlini S. & Lacchi L. (Eds.), Geologyof Italy, Società Geologica Italiana, Roma, 232 pp.

PICCARDI L. (1998) - Cinematica attuale, comportamento sismico e si-smologia storica della faglia di Monte Sant’Angelo (Gargano, Italia):la possibile rottura superficiale del «leggendario» terremoto del 493d.C. Geografia Fisica Dinamica Quaternaria, 21, 155-166.

PICCARDI L. (2005) - Paleoseismic evidence of legendary earthquakes.The apparition of Archangel Michael at Monte Sant’Angelo (Italy).Tectonophysics, 408, 113-128.

RAMBERG H. (1981) - Gravity, deformation Earth’s crust. AcademicPress, 452 pp., London.

RICHARD P.D., BALLARD J.F., COLLETTA B. & COBBOLD P.R. (1989)- Naissance et evolution de failles au-dessus d’un decrochementde socle; modelisation analogique et tomographie. ComptesRendus de l’Academie des Sciences, Serie 2, Mecanique, Physi-que, Chimie, Sciences de l’Univers, Sciences de la Terre, 309,2111-2118.

RICHARD P.D. & COBBOLD P. (1990) - Experimental insights into par-titioning of fault motions in continental convergent wrench zones.Annales Tectonicae, Special Issue, 4 (2), 35-44.

RICHARD P.D & KRANTZ R.W. (1991) - Experiments on fault reactiva-tion in strike-slip mode. Tectonophysics, 188, 117-131.

RICHARD P.D., NAYLOR M.A. & KOOPMAN A. (1995) - Experimentalmodels of strike slip tectonics. Petroleum Geoscience, 1, 71-80.

RIDENTE D. & TRINCARDI F. (2006) - Active foreland deformation evi-denced by shallow folds and faults affecting late Quaternary shelf-slope deposits (Adriatic Sea, Italy). Basin Research, 18 (2), 171-188. doi:10.1111/j.1365-2117.2006.00289.x

RIDENTE D., FRACASSI U., DI BUCCI D., TRINCARDI F. & VALENSISE

G. (2008) - Middle Pleistocene to Holocene activity of the GondolaFault Zone (Southern Adriatic Foreland): deformation of a regio-nal shear zone and seismotectonic implications. Tectonophysics,453 (1-4), 110-121. doi:10.1016/j.tecto.2007.05.009.

SASSI W., COLLETTA B., BALÉ P. & PAQUEREAU T. (1993) - Modellingof structural complexity in sedimentary basins: the role of pre-exi-sting faults in thrust tectonics. Tectonophysics, 226, 97-112.

SCHELLART W.P. (2000) - Shear test results for cohesion and frictioncoefficients for different granular materials: scaling implicationsfor their usage in analogue modelling. Tectonophysics, 324, 1-16.

SCHOLZ C.H. (1990) - The mechanics of earthquakes and faulting.Cambridge University Press, 439 p.

SCHÖPFER M.P.J. & STEYRER H.P. (2001) - Experimental modeling ofstrike-slip faults and the self-similar behavior. In: Koyi H.A. &Mancktelow N.S. (Eds.), Tectonic Modeling: A Volume in Honorof Hans Ramberg. Geological Society of America Memoir, 193,21-27.

SCHREURS G. et alii (2006) - Analogue benchmark of shortening andextension experiments. In: Buiter S.J.H. & Schreurs G. (Eds.),Analogue and Numerical Modelling of Crustal-Scale Processes.Geological Society of London, Special Publications, 253, 1-27.

SERPELLONI E., VANNUCCI G., PONDRELLI S., ARGNANI A., CASULA

G., ANZIDEI M., BALDI P. & GASPERINI P. (2007) - Kinematics ofthe Western Africa-Eurasia plate boundary from focal mechani-sms and GPS data. Geophysical Journal International, 169 (3),1180-1200. doi:10.1111/j.1365-246X.2007.03367.x.

SYLVESTER A.G. (1988) - Strike-slip faults. Geological Society ofAmerica Bulletin, 100, 1666-1703.

TCHALENKO J.S. (1970) - Similarities between shear zones of differentmagnitudes. Geological Society of America Bulletin, 81, 1625-1640.

TONDI E., PICCARDI L., CACON S., KONTNY B. & CELLO G. (2005) -Structural and time constraints for dextral shear along the sei-smogenic Mattinata fault (Gargano, southern Italy). Journal ofGeodynamics, 40, 134-152.

TURRINI C., RAVAGLIA A. & PEROTTI C.R. (2001) - Compressionalstructures in a multilayered mechanical stratigraphy: insightsfrom sandbox modelling with three-dimensional variations in ba-sal geometry and friction. In: Koyi H.A., Mancktelow N.S. (Eds.),Tectonic Modeling: A Volume in Honor of Hans Ramberg. Geo-logical Society of America Memoir, 193, 153-178.

UNITED STATES GEOLOGICAL SURVEY (1990) - Professional Paper1515: «The San Andreas Fault System, California». Wallace R.E.(ed.). United States Government Printing Office, Washington.

VALENSISE G., PANTOSTI D. & BASILI R. (2004) - Seismology and tec-tonic setting of the 2002 Molise, Italy earthquake. EarthquakesSpectra, 20, 24-37. doi:10.1193/1.1756136.

VALLÉE M. & DI LUCCIO F. (2005) - Source analysis of the 2002 Moli-se, southern Italy, twin earthquakes. Geophysical Research Let-ters, 32, 12, 4 pp.

VIOLA G., ODONNE F. & MANTCKELOV N.S. (2004) - Analogue model-ling of riverse fault reactivation in strike-slip and transpressive re-gimes: application to the Giudicarie fault system, Italian easternAlps. Journal of Structural Geology, 36, 401-418.

WILCOX R.E., HARDING T.P. & SEELY D.R. (1973) - Basic wrench tec-tonics. The American Association of Petroleum Geologists Bulle-tin, 57 (1), 74-96.

WINTER T. & TAPPONNIER P. (1991) - Extension majeure post-Jurassi-que et ante-Miocene dans le centre de l’Italie: données microtecto-niques. Bull. Soc. Géol., France, 162 (6), 1095-1108.


Received X Xxxxxxxx XXXX; revised version accepted X Xxxxxxxxx XXXX.

Date post:	25-Sep-2018
Category:	Documents
Upload:	vunhan
View:	212 times
Download:	0 times

Propagation of an inherited strike-slip fault through … et al., 2009... · mazioni a carico di un...

Documents