Insights on the seismogenic layer thickness from the upper crust structure of the Umbria-Marche Apennines (central Italy)

Transcription

1 TECTONICS, VOL. 27,, doi: /2007tc002134, 2008 Insights on the seismogenic layer thickness from the upper crust structure of the Umbria-Marche Apennines (central Italy) F. Mirabella, 1 M. Barchi, 1 A. Lupattelli, 1 E. Stucchi, 2 and M. G. Ciaccio 3 Received 20 March 2007; revised 14 November 2007; accepted 20 November 2007; published 14 February [1] We reconstruct the subsurface geology in a region of the northern Apennines (central Italy) where a protracted extensional sequence occurred in with maximum magnitude M = 6.0. Our study is mainly based on the interpretation of three reprocessed seismic reflection profiles crossing the epicentral area, which constrain the subsurface geometry to a depth of about 12 km where most of the shallow seismicity occurs. Comparing the subsurface setting with accurately determined earthquake locations, we find that the seismicity is located entirely within the sedimentary cover and does not penetrate the underlying basement. This is explained by considering that the sedimentary cover is rather thick and composed of relatively strong lithologies (platform carbonates and evaporites), while the upper part of the basement consists of weak phyllites and siliciclastic rocks. This weak horizon is also evidenced by the low-vp values measured in deep wells of the region. Its effect is to decouple the sedimentary cover from the crystalline basement, where only microseismicity occurs. Our study indicates that local structure and stratigraphy can significantly influence the distribution of seismicity within the upper crust, particularly in complex geological environments such as thrust-and-fold belts. Citation: Mirabella, F., M. Barchi, A. Lupattelli, E. Stucchi, and M. G. Ciaccio (2008), Insights on the seismogenic layer thickness from the upper crust structure of the Umbria-Marche Apennines (central Italy), Tectonics, 27,, doi: /2007tc Introduction [2] Geophysical data are commonly used to define the boundary between brittle and ductile behavior (the B/D transition) within the crust. Beneath the continents this transition is no deeper than about 50 km, but varies from 1 Geologia Strutturale e Geofisica, Earth Sciences Department, University of Perugia, Perugia, Italy. 2 Geophysics, Earth Sciences Department, Milano, Italy. 3 Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy. Copyright 2008 by the American Geophysical Union /08/2007TC a few km to tens of km depending on the geological environment [e.g., Watts and Burov, 2003]. [3] The mode of deformation in the upper and lower parts of the crust (brittle and ductile) depends on various physical parameters which control rock rheology: pressure, temperature, pore pressure content, the strength of the rocks, and internal friction. In the brittle part of the continental crust, above the B/D transition, deformation can be either seismic, microseismic or aseismic. Depending on the interaction of the above parameters, seismic deformations usually occur within the upper 20 km (the seismogenic layer), while at greater depths semibrittle deformations are common [e.g., Scholz, 1990]. [4] The depth of the seismogenic layer is defined by the cut-off of seismicity [e.g., Jackson and White, 1989]. While temperature is the most important factor controlling the thickness of the seismogenic layer, the composition and stratigraphy of the crust also play an important role. As observed by Sibson [1984], seismic deformations within the brittle crust can cluster near velocity boundaries owing to the contrasting elastic properties of the rocks [Faccenda et al., 2008]. The depth of the seismogenic layer can therefore be identified by lithological and/or structural discontinuities within the upper crust stratigraphy. [5] In this paper we explore the seismicity distribution and subsurface structure of the Umbria-Marche Apennines of Italy, where a protracted earthquake sequence occurred in on a set of SW-dipping normal faults (Figure 1a). The sequence began in September 1997 in the Colfiorito area (Figure 1b), and was characterized by the occurrence of six events with Mw > 5, the most energetic of which with Mw = 6.0 [Amato et al., 1998; Chiaraluce et al., 2003]. In April 1998 the sequence migrated to the NW (in the Gualdo Tadino area) with a Mw = 5.1 event [Ciaccio et al., 2005]. A local network was activated soon after the beginning of the sequence, recording thousands of events. Most of these were successively relocated using double difference techniques, providing a detailed and precise 3-D image of the activated fault system [Chiaraluce et al., 2003]. [6] We reconstruct the subsurface geology of this region by interpreting three reprocessed seismic reflection profiles. The investigated depths (up to 5 s two-way-time, corresponding to a depth of km) are comparable to the nucleation depths of the mainshocks and their aftershock sequences (<10 km). A recent and detailed (scale 1:10,000) map of the epicentral area [Regione Umbria, 2002] provided surface geology data supporting our geological interpretation of the seismic profiles. 1of15

2 Figure 1 2of15

3 Figure 2. Schematic seismic stratigraphy of (left) the Umbria-Marche region and (right) the corresponding main lithological units and interval velocities described in the text. [7] The availability of well-constrained earthquake and structural data in this region provides a good opportunity for discussing the relationship between geological structure and seismicity. One goal of this research is to highlight the role of lithological and structural boundaries in controlling the depth and thickness of the seismogenic layer within the upper continental crust. [8] After a brief review of the geological and seismotectonic setting, we describe the seismic profiles, discuss the proposed interpretation and draw three geological cross sections and one longitudinal section. We then plot the seismicity of the sequence, onto these sections and discuss its relationship with the subsurface structure. 2. Tectonic Setting [9] The Umbria-Marche Apennines are an arc-shaped, NE-verging, thrust-and-fold belt corresponding to the easternmost part of the northern Apennines. A comprehensive review of the regional geology has recently been provided by Barchi et al. [2001]. [10] The thrust-and-fold belt involves the rocks of the sedimentary cover, which consists of three major lithological groups (Figure 2). In order from top to bottom, these are turbidites (Miocene, up to 3000 m thick), made from alternating layers of sandstones and marls; carbonates (Jurassic-Oligocene, about 2000 m thick), consisting of an early Jurassic carbonate platform (Calcare Massiccio Fm.), overlain by pelagic limestones with subordinated marly levels; and evaporites (late Triassic, m thick), made from alternating layers of anhydrites and dolomites (Anidriti di Burano Fm. [Martinis and Pieri, 1964]). [11] An important regional décollement separates these units from the underlying basement (in the sense of Bally et al. [1986]). The basement never crops out, is penetrated only by a few deep wells, and shows an accentuated lithological variability. Beneath the evaporites, some wells (e.g., S.Donato 1 [Anelli et al., 1994] and Perugia 2 Figure 1. (a) Geological sketch of the Umbria-Marche Apennines between Cittá di Castello to the NW and Norcia to the SE. The main compressional and extensional structures (ATF and UFS) of the area are drawn together with the historical seismicity (Me > 5.5) and the instrumental seismicity (M > 5.0). (b) Geological sketch of the Umbria-Marche Apennines between Gualdo Tadino to the NW and Visso to the SE representing the Inner and Outer ridges of the region separated by the Matelica-Camerino (MC) synclinorium [Scarsella, 1951]. The figure represents the main compressional and extensional structures, the mainshocks of the area in the last 20 years and the earthquake sequence. F1, F2, F3, and F4 are the main normal fault segments. The location of the reprocessed seismic lines (S1, S2, and S3) and of the integrated geological sections (G1, G2, and G3) are also drawn for further reference. 3of15

4 [Ghelardoni, 1962]) have drilled into the upper part of the basement (top of the basement), finding phyllites from the Upper Paleozoic age. Other wells (e.g., Alessandra 1, Puglia1 [Bally et al., 1986; Patacca and Scandone, 2001]) have penetrated siliciclastic rocks (mainly sandstones and slightly metamorphosed pelites) of the Permian and early Triassic lying beneath the Middle Triassic Tuscan Verrucano Group. Upper Paleozoic/lower Triassic rocks similar to those drilled in the Umbria-Marche wells are also exposed in Tuscany [Lazzarotto et al., 2003]. Despite their lithological variability, in situ measurements have shown that all these rocks possess a Vp between 4.8 and 5.2 km/s, significantly lower than that of the evaporites above (Vp > 6.0 km/s). This sharp decrease in Vp of the top of the basement rocks generates a characteristic group of highamplitude reflections that can easily be followed in the seismic profiles (Figure 2). [12] The thickness of the low-velocity layer, however, is not well constrained by available data. According to seismic refraction surveys [Ponziani et al., 1995; De Franco et al., 1998] it is at least 2 km thick, even though the wells have penetrated a maximum thickness of 1500 m [Anelli et al., 1994]. The genuine crystalline basement of the Adria plate lies somewhere below this depth, and has never been reached by any well. Its lithological and geophysical characteristics can only be inferred by analogy with the westernmost Tuscan region, where a complex assemblage of michascists and gneiss was drilled in the Larderello geothermal area [Batini et al., 1983]. Refraction DSS data indicate that below the Umbria-Marche region the crystalline basement is characterized by Vp values of about 6 km/s [Ponziani et al., 1995; De Franco et al., 1998]. [13] The deep structural style of the Umbria-Marche Apennines has been debated for a long time. Bally et al. [1986], through an extensive study of seismic reflection data, first applied to this region the concepts of thin-skinned tectonics (where the evaporites were considered the main basal detachment). Later interpretations of seismic reflection profiles showed that at least the upper part of the basement is also involved in the main thrust sheets [Barchi, 1991; Sage et al., 1991; Barchi et al., 1998a]. A thickskinned model based on the inversion of Permo-Triassic basins was later proposed by Coward et al. [1999] and extended to other sectors of the Apennines [Butler et al., 2004]. [14] The late Miocene, arc-shaped compressional structures are obliquely dissected by a set of NW SE trending normal faults, which have been active at least since the early Pleistocene. These faults, which border intermontane continental basins from Sansepolcro to Norcia, are here named the Umbria Fault System (UFS, Figure 1a). The present-day activity of the UFS is addressed by a large body of geomorphological, geological and geophysical data [Lavecchia et al., 1994; Barchi et al., 2000; Chiaraluce et al., 2005, and references therein]. [15] Geodetic data indicate a relatively low extensional rate of 2.5 mm/a [Hunstad et al., 2003] along a SW NE axis, perpendicular to the mean trend of the UFS. These normal faults are responsible for historical events and the instrumental seismicity currently recorded in this region. Several extensional earthquakes of moderate magnitude (5 < M < 6) have occurred over the last 25 years, in 1979 (Norcia, Ms = 5.9 [Deschamps et al., 1984]), 1984 (Gubbio, Ms = 5.2 [Haessler et al., 1988]), and (the Colfiorito and Gualdo Tadino events, with several mainshocks with M > 5 [e.g., Amato et al., 1998; Ciaccio et al., 2005]). [16] The comparison between geological data (the kinematics of the Quaternary faults) and seismological data (the focal mechanisms of the earthquakes) has shown that the present-day extension affected the region since the Early Pleistocene with the same SW NE orientation of extension [Lavecchia et al., 1994; Calamita et al., 2000; Mirabella and Pucci, 2002; Chiaraluce et al., 2005]. [17] Seismic profiles in the area between Gubbio and Perugia have shown that the SW-dipping Gubbio normal fault, in the NW portion of the UFS, is antithetic to an ENEdipping low-angle normal fault bordering the northern Tiber Valley (the Altotiberina Fault [Barchi et al., 1998b; Boncio et al., 2000]; see Figure 1a) and releasing abundant microseismicity [Boncio et al., 1998; Chiaraluce et al., 2007, and references therein]. 3. Seismic Reflection Data [18] The study area was explored in the 1980s by Agip (the Italian oil company, presently ENI E&P), who acquired seismic sections and drilled deep wells. In recent years some of these data have been made available for scientific studies. Extensive interpretations of seismic data started during the CROP-03 (deep crust seismic reflection acquisition) project [Barchi et al., 1998a]. Later, seismic profiles crossing the area between Perugia and Gubbio were studied in detail. These researches highlighted the subsurface geometries of the main extensional structures, such as the Alto Tiberina fault and the antithetic Gubbio normal fault [Barchi et al., 1999; Collettini et al., 2000; Pauselli et al., 2002; Mirabella et al., 2004]. [19] Here we consider a set of three seismic reflection profiles, the only seismic data available on the epicentral area. From north to south, the profiles are named S1, S2, and S3; their locations are shown in Figure 1b. These profiles were acquired in the early 1980s using the Vibroseis technique. The data were collected along a crooked trace (Figure 1b), so their general quality is not very good. It is common knowledge that seismic exploration in the Apennines is hampered by several factors: the complex tectonic setting; outcropping lithologies, especially fractured carbonate rocks; and poor accessibility [Mazzotti et al., 2000]. Low signal penetration characterizes these data, particularly in the mountainous areas of the Inner Ridge where the carbonates are exposed at the surface. Moreover, the original processing was carried out in the 1980s, with relatively deep oil targets as main objectives. [20] In order to mitigate these negative effects, we reprocessed the portions of the profiles crossing the epicen- 4of15

5 tral area. Our aim was to improve the signal-to-noise ratio of the whole section, paying attention both to deep and shallow events. We obtained better imaging of those structures lying close to the surface, which can more easily be connected with the surface geology. At depth, the continuity of the top of the basement reflector was significantly improved. A detailed description of the reprocessing procedure and its goals is given by Stucchi et al. [2006]. [21] This paper presents the first geologically coherent interpretation of the three re-processed profiles (Figures 3, 4, and 5). Previous interpretations of the original profiles were published for sections S2 [Mirabella and Pucci, 2002] and S3 Barchi [1991]. A preliminary interpretation of the reprocessed S1 profile was published by Ciaccio et al. [2005]. [22] Even after reprocessing, the nature of exposed rocks along the trace significantly influences the quality of the seismic profiles. Where the outcropping rocks consist of marls and sandstones, the signal penetrates to great depths and reveals the geometry of the subsurface carbonates structures. This is the case for the western and eastern portions of profiles S1 (Figure 3) and S2 (Figure 4), which cross Miocene turbidites exposed in the Umbria pre-apennine and Matelica-Camerino synclinorium, respectively. On the other hand, when these sections cross the highly deformed and fractured limestone of the Inner Ridge, signal penetration is low and the subsurface images are of lower quality. This is the case for the central parts of sections S1 and S2 (Figures 3 and 4), as well as the entire length of section S3 (Figure 5) which crosses only carbonatic rocks. [23] Within these limitations, the seismic profiles used in this work show the subsurface structures to a maximum depth of 5 s (twt), corresponding to km. The correspondence between the main lithological units of the regional stratigraphy, and the seismic stratigraphy (Figure 2) recognizable in the profiles, was established on the basis of previous experience in the same region and through the calibration of the deep wells [e.g., Bally et al., 1986; Barchi et al., 1998a; Mirabella et al., 2004]. From top to bottom: the turbidites correspond to scattered, low-amplitude and scarcely continuous reflections, often interested by lowwavelength deformations; the carbonates show some more continuous, parallel reflectors: among them, the prominent high-amplitude reflectors generated by the Cretaceous, marly horizon of the Marne a Fucoidi Fm.; the evaporites are generally characterized by a light and transparent facies; the top of the basement is signed by a group of highamplitude, obliquely stratified reflections, that can be easily followed throughout the region. [24] The northernmost seismic profile, S1 (Figure 3; see trace in Figure 1b), is about 30 km long and crosses the town of Gualdo Tadino. The western part of the profile crosses the Umbria pre-apennines, where the turbidites are exposed and the subsurface compressional structures involving the carbonates are relatively simple. This portion of the subsurface is therefore well imaged by the seismic data. We can recognize an ENE-verging anticline, with its western limb downthrown by a SW-dipping normal fault. This feature represents a continuation of the Gubbio structure, which is exposed at the surface about 15 km to the NW [Mirabella et al., 2004]. The central part of the S1 section crosses the carbonates of the Inner Ridge, where only sparse reflections of the Marne a Fucoidi Fm are present. In the eastern part, below the Matelica-Camerino synclinorium, we observe a wide anticline sustained by a SW-dipping thrust. Below there is evidence for a further west-dipping reflection, representing a deeper tectonic unit involving both carbonates and evaporites. The highly reflective horizons of the top of the basement are continuously imaged throughout the S1 section. In the western part, almost flat basement reflections, pertaining to two imbricated thrust sheets, can be easily recognized at about 2.2 and 2.7 s, respectively. In the eastern part, SW-dipping basement reflections are nicely depicted, which can be followed also below the central part of the section, imaging a regular, SW-dipping monocline, deepening from 2.5 down to 3.5 s from east to west. [25] The central seismic profile, S2 (Figure 4, trace in Figure 1b), is about 35 km long and runs from Assisi to Camerino passing to the north of the village of Colfiorito. To the west it crosses the Mt. Subasio anticline, where carbonates emerge. East of this structure Miocene turbidites are exposed at the core of a relatively narrow syncline, corresponding to the Topino river valley. Beneath this zone, the prominent east-dipping reflector of the Marne a Fucoidi horizon generates a clear subsurface image of the Mt. Subasio anticline s eastern limb. Below Mt. Subasio, the flat reflections at s and s represent submerged carbonates and evaporites respectively. As for the central and eastern parts of the profile, the trace crosses first the Inner Ridge, then the Matelica-Camerino synclinorium. Below both regions SW-dipping reflections corresponding to submerged carbonate thrust sheets are clearly visible: for example, west of Colfiorito (from 0.5 to 2 s), east of Mt. Pennino (from 2 to 2.5 s), and below the Matelica-Camerino synclinorium at depths of s and s. These data have proved particularly useful in constraining the geometry of imbricated thrust sheets in the depth-converted sections. The shallow reflectors can be easily connected to major thrusts emerging at the surface. In the central portion of the profile, one of the main normal faults activated during the crisis (the Mt. LeScalette Mt.Pennino normal fault [Mirabella and Pucci, 2002; Chiaraluce et al., 2005]) can be also recognized. Owing to the crooked trajectory of the profile, it sometimes crosses the geological structures along their strike. This is the case between Mt. Faeto and Colfiorito (see Figure 1b), where the section runs NW SE along the normal fault s strike direction [Stucchi et al., 2006]. At greater depths there is a high-amplitude group of signals about 0.5 s thick, which image the top of the basement. This section shows that the basement consists of three major slices, separated by major west-dipping thrust faults (th1 and th2 in Figure 4). The upper tectonic unit, imaged below Mt. Subasio, is either flat or gently eastdipping from 2.5 to 3.0 s. A lower west-dipping slice can be recognized between 3.0 and 3.5 s. The lowest tectonic unit is a continuous basement monocline dipping from 2.6 s down to 3.5 s (east to west), below the Inner ridge and the Matelica-Camerino Synclinorium. 5of15

6 Figure 3. (top) Reprocessed S1 seismic profile (trace in Figure 1b) crossing the Inner ridge of the Umbria-Marche Apennines and (bottom) geological interpretation. Here tmf, tbas, and MA refer to top Marne a Fucoidi, top of the basement, and Miocene turbidites reflections, respectively; ATF represents the AltoTiberina normal fault reflector; GTF is the Gualdo Tadino normal fault [after Ciaccio et al., 2005]; and th1 and th2 are the main thrusts involving the basement. 6of15

7 Figure 4. (top) Reprocessed S2 seismic profile (trace in Figure 1b) crossing the Inner ridge of the Umbria-Marche Apennines and (bottom) geological interpretation. Here tmf, tbas, and MA refer to top Marne a Fucoidi, top basement, and Miocene turbidites reflections, respectively; and th1 and th2 are the main thrusts involving the basement (see text for details). 7of15

8 Figure 5. (top) Reprocessed S3 seismic profile (trace in Figure 1b) crossing the Inner ridge of the Umbria-Marche Apennines and (bottom) geological interpretation. Here tmf and tbas refer to top Marne a Fucoidi and top basement, respectively. 8of15

9 [26] The southernmost section, S3 (Figure 5, trace in Figure 1b), starts 8 km east of the town of Foligno. From there it crosses the Inner Ridge to reach the southern termination of the Matelica-Camerino synclinorium, near Visso. Close to the surface, some shallow reflectors can be associated with the major thrusts emerging at the surface. At intermediate depths (from 1 to 2 s), only sparse SW-dipping reflections are evident in the western and central parts of the section. At great depths (from 1.5 to 2.5 s), the reflections in the eastern part of the section are interpreted as pertaining to carbonate thrust sheets. Finally, an almost flat, continuous reflection is located at s in the central part of the section. This signal may correspond to the top of the evaporites. The reprocessing enhanced the continuity of the deeper and stronger reflections, which image the top of the basement which, in this section, deepens westward from 3.2 s to about 4 s. [27] To summarize, these three seismic profiles depict some major structural features of the region. The deepest structure is a continuous, gently west-dipping basement monocline which extends below the Inner Ridge of the Umbria-Marche Apennines. The depth of the top of the basement ranges from 2.5 to 3 s in the eastern part of the region, below the Matelica-Camerino synclinorium, and from 3.5 to 4 s in the western part. Comparing the three sections also reveals that the top of the basement slightly deepens from NNW to SSE. Seismic profiles S1 and S2 show other structures below the Umbria pre-apennines, where Miocene turbidites are exposed. Here the top of the basement is much shallower (2 to 3 s), and displaced by at least two west-dipping thrust faults. [28] The internal structure of the sedimentary cover is not completely defined by the seismic profiles, but sparse reflections clearly indicate that it consists of imbricated, west-dipping thrust sheets. These can be connected with the folds and thrusts exposed at the surface. [29] The basement structure, consisting of two thrusts, is much simpler than the shallower structure, involving the carbonates and evaporites thrust sheets. This interpretation implies that the basement thrusts branch into multiple shallower structures to form the complex fold-and-thrust belt observed at the surface. This result confirms the role played by the evaporites as the major décollement of the Umbria- Marche region [Baldacci et al., 1967; Bally et al., 1986]. [30] The seismic expression of the Quaternary normal faults, which dissect the previously formed fold-and-thrust belt, is actually very weak. This can be explained by the fact that the throws of the normal faults are relatively small (less than 500 m [Mirabella et al., 2005]) compared to the shortening accommodated by the major thrusts, which usually exceeds 10 km. As a consequence, the seismogenic normal faults separate essentially similar lithologies, and generate no strong reflections. 4. Geological Sections [31] Starting from the above interpretation of the seismic profiles, we drew three geological cross sections named G1, G2, and G3 from north to south (Figure 6, traces in Figure 1b). The traces of the geological cross-sections are straight and fit the crooked path of the seismic profiles. Sections G1 and G2 trend about N80, while G3 trends roughly E W. They are thus drawn roughly perpendicular to the arc-shaped compressional structures and slightly oblique to the NW SE striking extensional structures (Figure 1b). [32] The geological sections were drawn using the following procedure: (1) Detailed geological sections of the surface were drawn using data derived from recent, detailed (1:10,000 scale) geological surveys of the study area [Regione Umbria, 2002]. (2) The signals observed along the seismic profiles were projected onto these sections, then vertically depth-converted using average interval velocities derived from in situ measurements made in deep wells. The interval velocity data necessary for depth conversion were previously analyzed and compiled [Bally et al., 1986; Barchi et al., 1998a]. (3) The surface geological sections were integrated with the subsurface structures identified in the seismic data. This step relied partly on previous experience with the structural style of the region, as derived from the interpretation of the deep CROP-03 profile and many other commercial profiles [Barchi, 1991; Barchi et al., 1998a, 1999; Mirabella et al., 2004]. (4) Each section was 2-D balanced with the software 2D-Move 2, in order to assure the kinematic consistency of the structures and in particular of the subdued carbonatic sheets. [33] The three sections shown in Figure 6 (G1, G2, and G3) show the compressional and extensional structures of the Inner Ridge of the Umbria-Marche Apennines. The sections are spaced about 15 km apart (Figure 1b), allowing us to visualize along-strike variations of the geometry of the geological structures. [34] The Inner Ridge widens from north to south, but is still being formed by two main thrusts (labeled th1 and th2) involving the top of the basement which are recognizable in sections G1 and G2. The deep thrusts branch out as they approach the surface, forming a complex pattern of shorterwavelength anticlines involving both carbonates and evaporites. [35] The geological sections also show the depth and geometry of the basement as calibrated by the seismic reflection data. On the basis of our interpretation, thickand thin-skinned tectonics coexist in this region. The basement is involved in the compressional structures in the western part of the sections, while below the Inner Ridge, the basement is a west-dipping monocline, deepening from about 6 to about 9 km. Figure 6. Integrated geological cross sections (G1, G2, G3, and G4, traces in Figure 1b) obtained integrating the seismic sections S1, S2, and S3, surface geology, and regional knowledge about the deformation style. G1, G2, and G3 are transversal (SW NE) sections, and G4 is a longitudinal section (trending NW SE), mainly based on the transversal sections and surface geology. Here th1 and th2 are the main thrusts involving the basement. See text for details. 9of15

10 Figure 6 10 of 15

11 [36] As already discussed, the seismic profiles do not provide relevant information about the geometry of the active normal faults. In the sections of Figure 6, we have traced the normal faults based on two factors: (1) the positions, attitudes and throws observed at the surface during a detailed mapping of the study area [Regione Umbria, 2002; Mirabella et al., 2005], and (2) the geometries of activated fault segments imaged at depth (1 8 km) by the seismological data [Chiaraluce et al., 2005]. [37] In order to image along-strike structural variations and determine their relationship to seismicity, we also draw a longitudinal section (G4 in Figure 6) trending about N140 (see trace in Figure 1b) parallel to the normal faults strike. The trace of the longitudinal section connects the mainshocks of the seismic sequence. This section was built up using mainly the surface data provided by field mapping. In the absence of any tie seismic profile, the geometry at depth is constrained only by the points of intersection with the previously described cross sections (G1, G2, G3). The three cross sections are spaced by only 15 km and it is reasonable to assume that the geometry of the thrust sheets does not vary significantly along strike over such a small interval. Section G4 shows a regional deepening of the basement monocline toward the SE, and indicates that the seismic sequence is located within the sedimentary cover, above the top of the basement. 5. Relationships Between Subsurface Structure and Seismicity [38] The trace of geological cross-section G1 (Figure 1b) runs about 3 km north of the Gualdo Tadino (1998) mainshock. Section G2 runs 5 km north of Colfiorito, where the most energetic event of the sequence occurred. Section G3 runs about 3 km north of the Sellano mainshock. [39] In order to analyze the relationships between the distribution of the seismicity and the subsurface structure, we projected onto each section the hypocenters of the closest mainshock and of the aftershocks, located in a strip of 2 km from the trace of the sections (Figure 6). The selected hypocenters were projected along the mean strike of the activated faults, i.e., along N140 [Chiaraluce et al., 2003, 2005]. Finally, we projected all the selected seismicity (both mainshocks and aftershocks) onto the longitudinal section G4, perpendicularly to the section itself (Figure 6). [40] The location of the seismic events is very well constrained. The seismological data of Colfiorito and Sellano (1650 events) were relocated using the double difference technique [Waldhauser and Ellsworth, 2000] with formal errors of 70, 85 and 120 m, respectively, in latitude, longitude and depth. Consequently, structural details with dimensions of about 100 m can be reliably interpreted [Chiaraluce et al., 2003]. The Gualdo Tadino data (216 events) were also relocated very precisely, with mean errors less than 0.3 km in all directions [Ciaccio et al., 2005]. [41] Our sections show that seismicity associated with the sequence is completely located within the sedimentary cover, which consists of carbonates and evaporites; it does not penetrate into the basement rocks. In other words, the seismicity cut-off seems to be related to the depth of the top of the basement. The mainshocks of the sequence (M > 5.0) are located within Triassic evaporites, which consist of anhydrites and dolomites in alternating layers. Along the longitudinal section G4, which runs from NW to SE, we see that the seismicity is located above basement layers which are associated with progressively deeper tectonic units: the hanging wall of th1, the hanging wall of th2, and the footwall of th2. This is due to the attitude of the active normal fault system, which obliquely dissects the arc-shaped thrust belt (Figure 1b). [42] These features are not peculiar to the Colfiorito area, but can be extended to other seismogenic normal faults of the Umbria region (UFS) by considering the available data on instrumental seismicity and the subsurface structure. In the last 25 years, two other moderate extensional earthquakes pertaining to the UFS have occurred in this region (Figure 1a). The first occurred in 1979 in the Norcia area (Ms = 5.9), about 40 km SE of Colfiorito [Deschamps et al., 1984]; the second occurred in 1984 in the Gubbio area (Ms = 5.2), about 40 km NW of Colfiorito [Haessler et al., 1988; Collettini et al., 2003]. [43] In the Gubbio area, a set of high-quality seismic sections indicates a similar correspondence between the seismicity cut-off and the top of the basement, which there lies at a depth of 6 7 km [Mirabella, 2002; Collettini et al., 2003]. In the Norcia area, available seismological data indicate that the seismicity cut-off is about 12 km below the surface. No seismic profile is available for this region to corroborate this depth. However, Bally et al. [1986] hypothesized that the depth of the top of the basement is about 15 km, in accordance with regional geological data suggesting a steady southward increase in the top of the basement depth. A comparison of the maximum focal depths found in Gubbio (6 7 km), Colfiorito (8 9 km), and Norcia (about 12 km) does suggest a NW SE (i.e., along-strike) deepening of the seismogenic layer. [44] To summarize, available data on the moderate extensional seismicity of the Umbria-Marche region provide evidence for a lithological control on the thickness of the seismogenic layer [Barchi, 2002]. In fact, the seismicity is located within the sedimentary cover; it does not penetrate the underlying basement rocks. [45] This behavior is quite different from that seen in most seismogenic normal faults worldwide, which nucleate within the basement and crosscut the entire brittle upper crust [Jackson and White, 1989]. This difference can be related to some peculiarities of the cover and of the basement rocks of the Umbria-Marche region. [46] The first peculiarity is that the sedimentary cover is greatly thickened by the horizontal shortening produced by thrust tectonics. As the sections of Figure 6 show, below the Inner Ridge the present-day depth of the top of the basement ranges from 6 to 9 km, while the original thickness of the carbonates and evaporites is less than 4 km. [47] The second peculiarities is the presence of a low-vp horizon (top of the basement) which is interpreted as a weak layer. This layer consists of both clastic rocks (sandstone and pelites) and metamorphic rocks (phyllites), and is 11 of 15

12 located above the basement. The presence of this weak layer, which is about 2 km thick, partially decouples the sedimentary cover from the crystalline basement. [48] Unfortunately, little is known about the lithology and mechanical properties of this weak layer. The lithology can be extrapolated from corresponding outcrops in Tuscany [Lazzarotto et al., 2003] west of the study area, and from a few deep wells in the Umbria-Marche Apennines [Ghelardoni, 1962; Barchi et al., 1998a]. The mechanical properties can be inferred from the Vp values, which have been measured in the same wells and in a few laboratory tests [Burlini and Tancredi, 1998]. These data indicate that the top of the basement is a low-velocity layer (LVL), characterized by a Vp change from more than 6.0 km/s within the evaporites above to less than 5.0 km/s below. This inversion suggests a corresponding decrease in the elastic parameters of these rocks [e.g., Jaeger and Cook, 1969]. [49] The along-strike (NW SE) deepening of the seismogenic layer has been observed by other authors, who have suggested a structural control [Boncio and Lavecchia, 2000]. This model is based on the fact that in the region lying between Gubbio and Perugia (Figure 1a), seismic reflection profiles show that the SW-dipping Gubbio normal fault is antithetic to a major, ENE-dipping, low-angle normal fault named the Altotiberina Fault (ATF) [Boncio et al., 1998; Barchi et al., 1999; Collettini et al., 2003; Mirabella et al., 2004]. Boncio and Lavecchia [2000] extended this observation farther to the SE, reaching the Colfiorito and Norcia fault systems. According to this idea, all of the SW-dipping normal faults in the UFS, which produce the moderate seismicity of the region, may be rooted in this major ENE-dipping detachment (the ATF). Thus it is the intersection between SWdipping faults and the ATF, which occurs at progressively greater depths from NW (Gubbio) to SE (Norcia), that induces a deepening of the associated seismicity. [50] In fact, we do not recognize the ATF trace in the seismic profiles S2 and S3 presented in this study. However, we cannot exclude the possibility that the ENE-dipping detachment is located at a greater depth, below the resolution of our profiles. The ATF trace in the western part of section G1 is calibrated through a network of other seismic profiles, located immediately to the north [Mirabella et al., 2004]. Besides, considering the average strike of the ATF, as reconstructed below the Gubbio area [Barchi et al., 1999; Mirabella, 2002; Chiaraluce et al., 2007], even if the ATF were present below Colfiorito, its intersection with the seismogenic SW-dipping faults should occur at a depth greater than 12 km. This is considerably deeper than the observed seismicity cut-off. To summarize, in our opinion the presence of the ATF (or a similar ENE-dipping detachment) in the Colfiorito and Norcia areas is still an open and reasonable possibility, but is not supported by the available seismic data. 6. Discussion 6.1. Upper Crust Structure and Seismogenic Layer [51] Our data show that the seismicity related to moderate earthquakes in the study area is confined to the uppermost km, within the sedimentary cover. In the same region the cutoff of background seismicity registered by the Italian national network lies at a depth of km [Chiarabba et al., 2005], roughly corresponding to the B/D transition as evaluated by Pauselli and Federico [2002] using heat flow data. [52] In our interpretation, the brittle crust is partitioned into two seismogenic layers. A low-velocity horizon, corresponding to the shallower part of the basement, decouples the overlying sedimentary cover (where moderate seismicity is produced) from the underlying crystalline basement (where only microseismicity occurs). [53] The thickness of the seismogenic layer is generally thought to influence the maximum length of the active faults, and consequently the maximum expected magnitude [e.g., Jackson and White, 1989; Wells and Coppersmith, 1994]. Since in our view the depth of the top of the basement controls the maximum depth of the major earthquakes, the precise mapping of this discontinuity may provide insight into the maximum expected magnitude. In the northern Apennines, available data seem to confirm this speculation. The data show a general decrease in the magnitude of mainshocks toward the NW, associated with a similar decrease of the depth of the top of the basement in the same direction Mechanical Behavior of the Evaporites [54] Our data indicate that the mainshocks of Umbria- Marche seismicity occur at a depth of about 6 km, within the evaporites (Figure 6). This fact raises a question as to the mechanical properties of this formation, which has traditionally been considered a weak horizon corresponding to the basal décollement of the Umbria-Marche fold-andthrust belt. Its thickness greatly contribute to the bulk properties of the sedimentary cover. [55] In the Umbria-Marche Apennines this formation only crops out in the area of Mt. Malbe (a few kilometers NW of Perugia), where it is characterized by alternating layers of anhydrites (sometimes rehydrated to gypsum) and dolomites in similar proportions [Ciarapica and Passeri, 1976, 1980]. The same lithological assembly has been drilled in deep wells of the region (Burano1 [Martinis and Pieri, 1964; Perugia2 [Ghelardoni, 1962; S.Donato1 and M. Civitello [Anelli et al., 1994]). Vp values in the range of km/s (Figure 2) have been measured in the deep wells [Bally et al., 1986; Barchi et al., 1998a]. Similarly high Vp values have resulted from DSS seismic refraction experiments [Ponziani et al., 1995; De Franco et al., 1998], as well as from shallow tomography based on the seismic record of the Colfiorito area [Michelini et al., 2000; Chiarabba and Amato, 2003]. [56] Only a few laboratory measurements have tested the elastic properties of anhydrites and dolomites [Handin and Hager, 1957; Mogi, 1971; Mueller and Siemes, 1974; Bell, 1994; Mirabella, 2002], and none of these have considered the effect of temperature. The available data indicate that anhydrites and dolomites deform in a rather different manner: anhydrites are brittle at confining pressures less than 100 MPa, while dolomites maintain a brittle behavior at pressures greater than 200 MPa. This agrees with the brittle/ductile transition predicted by Paterson [1978] for 12 of 15

13 dolomites. Moreover, at similar confining pressures dolomites are stronger than anhydrites. At a confining pressure of 200 MPa, for example, the strength of dolomites is about 700 MPa while the strength of anhydrites is about MPa [Handin and Hager, 1957; Mueller and Siemes, 1974]. [57] To summarize, the available mechanical data indicate that dolomites should be still brittle, and harder than anhydrites, at a depth of about 6 km where the confining pressure is expected to be in the range of MPa. On this basis, we suggest that the dolomites are the mechanically dominant material of the evaporites. The dolomite layers could act as stress concentrators, inducing discrete faulting to propagate through the evaporites and also to affect the carbonates in the upper 2 3 km of the crust. The mode of failure of these rocks should also be investigated in the presence of overpressured fluids which could be of deep origin (i.e., mantle degassing) [Chiodini et al., 2000; Miller et al., 2004]. The presence of overpressured fluids, besides, could favor the reactivation of dipping faults, as indicated by the focal mechanisms and nodal planes of the mainshocks [Chiaraluce et al., 2003; Ciaccio et al., 2005]. 7. Conclusions [58] In this study we propose that the seismicity cut-off of the Umbria-Marche sequence is controlled by the top of the basement. The seismicity is confined to the sedimentary cover, which consists of carbonates and evaporites, and does not penetrate the underlying basement rocks. The mainshocks of the sequence (M > 5.0) are located within the Triassic evaporites, which consist of anhydrites and dolomites. In this region the sedimentary cover is thick and consists of relatively strong rocks, such as platform carbonates and dolomites. In contrast, the top of the basement is formed of relatively weak siliciclastic rocks and phyllites. [59] It would be interesting to discuss whether a similar situation occurs in other areas of the Apennines. In fact, the Umbria-Marche normal faults are one segment of a much longer extensional belt, extending from northern Tuscany to northern Calabria, which is characterized by both historical and instrumental seismicity [Pondrelli et al., 2006]. Deep wells drilled through the Apula platform of the southern Apennines foreland (Puglia1 and Gargano1) show that below the Late Triassic evaporites there is a layer of slightly metamorphosed siliciclastic rocks. The lithology, Vp, and stratigraphic position of these rocks are similar to those of the upper Umbria-Marche basement. [60] Some authors [Menardi Noguera and Rea, 2000; Patacca and Scandone, 2001] have drawn geological sections (mainly based on seismic data) in the region of the 1980 Irpinia earthquake (Ms = 6.9, estimated depth km [Deschamps and King, 1984]) in southern Italy. According to these sections, the bottom of the sedimentary cover (the Apula platform) is about 15 km deep [Improta et al., 2003]. If this reconstruction is correct, the 1980 Irpinia earthquake also occurred within the sedimentary cover (i.e., within the Apula platform carbonates or within the underlying evaporites). [61] Below the axial zone of the central Apennines, where the 1915 Avezzano earthquake occurred at an estimated depth of about 8 km (Ms = 7.0 [Boschi et al., 1997]), recently published data from the CROP-11 profile [Billi et al., 2006] have shown a deep reflector which is very similar to the top of the Umbria-Marche basement. This reflector is located at about 5 s twt, i.e., at a depth of about 13 km. Our study confirms that the rheology of the upper brittle crust can be influenced by its lithological composition and stratigraphy [Sibson, 1984; Jackson and White, 1989]. Most studies describe a quartz-dominated upper crust. This is not the case for some important geological environments such as fold-and-thrust belts, however, where a thick pile of sedimentary rocks can be involved in seismic faulting. More studies are needed to characterize the mechanical (and possibly seismogenic) behavior of such relatively strong sedimentary rocks (e.g., platform carbonates, anhydrites and dolomites), and also to examine their behavior in the presence of high-pressure fluids. [62] Acknowledgments. We thank Eni for providing the seismic sections. We thank C. Pauselli, P. Montone, and L. Valensise for fruitful discussions, and S. Bagh, S. Barba, and C. Collettini for reading an early version of the manuscript. We thank A. McGarr and an anonymous reviewer for their constructive and helpful advice. This work was supported by Miur 2001 and GNDT 2004 funding, through grants to M. Barchi. References Amato, A., et al. (1998), The 1997 Umbria-Marche, Italy earthquake sequence: A first look at the main shocks and aftershocks, Geophys. Res. Lett., 25, Anelli, L., M. Gorza, M. Pieri, and M. Riva (1994), Subsurface well data in the northern Apennines (Italy), Mem. Soc. Geol. Ital., 48, Baldacci, F., P. Elter, E. Giannini, G. Giglia, A. Lazzarotto, R. Nardi, and M. Tongiorgi (1967), Nuove osservazioni sul problema della Falda Toscana sulle interpretazioni dei Flysch arenacei di tipo Macigno dell Appennino Settentrionale, Mem. Soc. Geol. Ital., 6, Bally, A., L. Burbi, C. Cooper, and R. Ghelardoni (1986), Balanced sections and seismic reflection profiles across the central Apennines, Mem. Soc. Geol. Ital., 35, Barchi, M. (1991), Integration of a seismic profile with surface and subsurface geology in a cross-section through the Umbria-Marche apennines, Boll. Soc. Geol. Ital., 110, Barchi, M. (2002), Lithological and structural controls on the seismogenesis of the Umbria region: Observations from seismic reflection profiles, Boll. Soc. Geol. Ital., Vol. Spec. 1, Barchi, M., A. De Feyter, B. Magnani, G. Minelli, G. Pialli, and B. Sotera (1998a), The structural style of the Umbria-Marche fold and thrust belt, Mem. Soc. Geol. Ital., 52, Barchi, M., A. De Feyter, M. Magnani, G. Minelli, G. Pialli, and B. Sotera (1998b), Extensional tectonics in the northern Apennines (Italy): Evidence from the CROP03 deep seismic reflection line, Mem. Soc. Geol. Ital., 52, Barchi, M., S. Paolacci, C. Pauselli, G. Pialli, and S. Merlini (1999), Geometria delle deformazioni estensionali recenti nel bacino dell Alta Val Tiberina fra S.Giustino Umbro e Perugia: Evidenze geofisiche e considerazioni geologiche, Boll. Soc. Geol. Ital., 118, Barchi, M., F. Galadini, G. Lavecchia, P. Messina, A. Michetti, L. Peruzza, A. Pizzi, E. Tondi, and E. Vittori (2000), Sintesi delle conoscenze sulle faglie attive in Italia Centrale: Parametrizzazione ai fini della caratterizzazione della pericolositá sismica, Gruppo Naz. per la Difesa dai Terremoti, Rome. Barchi, M., A. Landuzzi, G. Minelli, and G. Pialli (2001), Outer northern Apennines, in Anatomy of an Orogen: The Apennines and Adjacent Mediterranean Basins, edited by G. Vai and I. Martini, pp , Kluwer Acad., Norwell, Mass. Batini, F., G. Bertini, G. Gianelli, and E. Pandeli (1983), Deep structure of the Larderello field: Contribution from recent geophysical and geological data, Mem. Soc. Geol. Ital., 25, of 15

14 Bell, F. (1994), A survey of the engineering properties of some anhydrite and gypsum from the north and midlands of England, Eng. Geol., 38, Billi, A., et al. (2006), First results from the CROP-11 deep seismic profile, central Apennines, Italy: Evidence of mid-crustal folding, J. Geol. Soc., 163, Boncio, P., and G. Lavecchia (2000), A structural model for active extension in central Italy, J. Geodyn., 29, Boncio, P., F. Ponziani, F. Brozzetti, M. Barchi, G. Lavecchia, and G. Pialli (1998), Seismicity and extensional tectonics in the northern Umbria- Marche Apennines, Mem. Soc. Geol. Ital., 52, Boncio, P., F. Brozzetti, and G. Lavecchia (2000), Architecture and seismotectonics of a regional lowangle normal fault zone in central Italy, Tectonics, 19(6), Boschi, E., E. Guidoboni, G. Ferrari, G. Valensise, and P. Gasperini (1997), CFTI, Catalogo dei Forti terremoti Italiani dal 461 a.c. al 1990, Ist. Naz. di Geofis., Bologna, Italy. Burlini, L., and S. Tancredi (1998), Experimental study of the seismic properties of Isola d Elba metamorphic basement (Tuscany, Italy), Mem. Soc. Geol. Ital., 52, Butler, R., et al. (2004), Applying thick-skinned tectonic models to the Apennine thrust belt of Italy Limitations and implications, in Thrust Tectonics and Hydrocarbon Systems, editedbyk. McClay, AAPG Mem., 82, Calamita, F., M. Coltorti, D. Piccinini, P. Pierantoni, A. Pizzi, M. Ripepe, V. Scisciani, and E. Turco (2000), Quaternary faults and seismicity in the Umbro-Marchean Apennines (central Italy): Evidence from the 1997 Colfiorito earthquake, J. Geodyn., 29, Chiarabba, C., and A. Amato (2003), Vp and Vp/Vs images in the Mw 6.0 Colfiorito fault region (central Italy): A contribution to the understanding of seismotectonic and seismogenic processes, J. Geophys. Res., 108(B5), 2248, doi: / 2001JB Chiarabba, C., L. Jovane, and R. Di Stefano (2005), A new view of the Italian seismicity using 20 years of instrumental recordings, Tectonophysics, 395, Chiaraluce, L., W. L. Ellsworth, C. Chiarabba, and M. Cocco (2003), Imaging the complexity of an active complex normal fault system: The 1997 Colfiorito (central Italy) case study, J. Geophys. Res., 108(B6), 2294, doi: /2002jb Chiaraluce, L., M. Barchi, C. Collettini, F. Mirabella, and S. Pucci (2005), Connecting seismically active normal faults with Quaternary geological structures in a complex extesional environment: The Colfiorito 1997 case history (northern Apennines, Italy), Tectonics, 24, TC1002, doi: /2004tc Chiaraluce, L., C. Chiarabba, C. Collettini, D. Piccinini, and M. Cocco (2007), Architecture and mechanics of an active low angle normal fault: The Alto Tiberina fault (northern Apennines, Italy) case study, J. Geophys. Res., 112, B10310, doi: / 2007JB Chiodini, G., F. Frondini, C. Cardellini, F. Parello, and L. Peruzzi (2000), Rate of diffuse carbon dioxide earth degassing estimated from carbon balance of regional aquifers: The case of central Apennines, Italy, J. Geophys. Res., 105(B4), Ciaccio, M., M. Barchi, C. Chiarabba, F. Mirabella, and E. Stucchi (2005), Seismological, geological and geophysical constraints for the Gualdo Tadino fault, Umbria-Marche Apennines (central Italy), Tectonophysics, 406, Ciarapica, G., and L. Passeri (1976), Deformazioni da fluidificazione ed evoluzione diagenetica della formazione evaporitica di burano, Boll. Soc. Geol. Ital., 95, Ciarapica, G., and L. Passeri (1980), Contributions on evaporites diagenesis, Geol. Mediterr., 7(1), Collettini, C., M. Barchi, C. Pauselli, C. Federico, and G. Pialli (2000), Seismic expression of active extensional faults in Northern Umbria (central Italy), J. Geodyn., 29, Collettini, C., M. Barchi, L. Chiaraluce, F. Mirabella, and S. Pucci (2003), The Gubbio fault: Can different methods give pictures of the same object?, J. Geodyn., 36, Coward, M., M. DeDonatis, S. Mazzoli, W. Paltrinieri, and F. Wezel (1999), Frontal part of the northern Apennines fold and thrust belt in the Romagna- Marche arc (Italy): Shallow and deep structural styles, Tectonics, 18(3), De Franco, R., F. Ponziani, G. Biella, G. Boniolo, G. Caielli, A. Corsi, M. Maistrello, and A. Morrone (1998), Dss-war experiment in support of the Crop03 project, Mem. Soc. Geol. Ital., 52, Deschamps, A., and G. King (1984), Aftershocks of the Campania-Lucania (Italy) earthquake of 23 November 1980, Bull. Seismol. Soc. Am., 74, Deschamps, A., G. Innaccone, and R. Scarpa (1984), The Umbrian earthquake (Italy) of 19 September 1979, Ann. Geophys., 2(1), Faccenda, M., G. Bressan, and L. Burlini (2008), Seismic properties of the upper crust in the central Friuli area (northeastern Italy) based on petrophysical data, Tectonophysics, 445, Ghelardoni, R. (1962), Stratigrafia e tettonica del Trias di M.Malbe presso Perugia, Boll. Soc. Geol. Ital., 81, Haessler, H., R. Gaulon, L. Rivera, R. Console, M. Frogneux, G. Gasparini, L. Martel, G. Patau, M. Siciliano, and A. Cisternas (1988), The Perugia (Italy) earthquake of 29 April 1984: A microearthquake survey, Bull. Seismol. Soc. Am., 78, Handin, J., and R. Hager (1957), Experimental deformation of sedimentary rocks under confining pressure: tests at room temperature on dry samples, Bull. Am. Assoc. Pet. Geol., 41(1), Hunstad, I., G. Selvaggi, N. D Agostino, P. Clarke, and M. Pierozzi (2003), Geodetic strain in peninsular Italy between 1875 and 2001, Geophys. Res. Lett., 30(4), 1181, doi: /2002gl Improta, L., M. Bonagura, P. Capuano, and G. Iannaccone (2003), An integrated geophysical investigation of the upper crust in the epicentral area of the 1980, Ms = 6.9, Irpinia earthquake (southern Italy), Tectonophysics, 361, Jackson, J., and N. White (1989), Normal faulting in the upper continental crust: Observations from regions of active extension, J. Struct. Geol., 11, Jaeger, J., and N. Cook (1969), Fundamentals of Rock Mechanics, Chapman and Hall, London. Lavecchia, G., F. Brozzetti, M. Barchi, J. Keller, and M. Menichetti (1994), Seismotectonic zoning in east-central Italy deduced from the analysis of the Neogene to present deformations and related stress fields, Geol. Soc. Am. Bull., 106, Lazzarotto, A., M. Aldinucci, S. Cirilli, A. Costantini, F. Decandia, E. Pandeli, F. Sandrelli, and A. Spina (2003), Stratigraphic correlation of the Upper Palaeozoic-Triassic successions in southern Tuscany, Italy, Boll. Soc. Geol. Ital., Vol. Spec. 2, Martinis, B., and M. Pieri (1964), Alcune notizie sulla formazione evaporitica dell Italia centrale e meridionale, Mem. Soc. Geol. Ital., 4, Mazzotti, A., E. Stucchi, G. Fradelizio, L. Zanzi, and P. Scandone (2000), Seismic exploration in complex terrains: A processing experience in the southern Apennines, Geophysics, 65, Menardi Noguera, A., and G. Rea (2000), Deep structure of the Campanian-Lucanian Arc (southern Appennine, Italy), Tectonophysics, 324, Michelini, A., D. Spallarossa, M. Cattaneo, A. Govoni, and A. Montanari (2000), The Umbria-Marche (Italy) earthquake sequence: Tomographic images obtained from data of the GNDT-SSN temporary network, J. Seismol., 4, Miller, S., C. Collettini, L. Chiaraluce, M. Cocco, M. Barchi, and B. Klaus (2004), Aftershocks driven by a high pressure CO 2 source at depth, Nature, 427, Mirabella, F. (2002), Seismogenesis of the Umbria- Marche region (Central Italy): Geometry and kinematics of the active faults and mechanical behaviour of the involved rocks, Ph.D. thesis, Univ. of Perugia, Perugia, Italy. Mirabella, F., and S. Pucci (2002), Integration of geological and geophysical data along a section crossing the region of the Umbria-Marche earthquake (Italy), Boll. Soc. Geol. Ital., Vol. Spec. 1, Mirabella, F., M. Ciaccio, M. Barchi, and S. Merlini (2004), The Gubbio normal fault (Central Italy): Geometry, displacement distribution and tectonic evolution, J. Struct. Geol., 26, Mirabella, F., V. Boccali, and M. Barchi (2005), Segmentation and interaction of normal faults within the Colfiorito fault system (Central Italy), in Deformation Mechanisms, Rheology and Tectonics: From Minerals to the Lithosphere, edited by D. Gapais, J. Brun, and P. Cobbold, Geol. Soc. Spec. Publ., 243, Mogi, K. (1971), Effect of the triaxial stress system on the failure of dolomite and limestone, Tectonophysics, 11, Mueller, P., and H. Siemes (1974), Festigkeit, verformbarkeit und gefuegeregelung von anhydrit Experimentelle stauchverformung unter manteldrucken bis 5 kbar bei temperaturen bis 300 c, Tectonophysics, 23, Patacca, E., and P. Scandone (2001), Late thrust propagation and sedimentary response in the thrust-beltforedeep system of the southern Apennines (Pliocene-Pleistocene), in Anatomy of an Orogen: The Apennines and Adjacent Mediterranean Basins, edited by G. Vai and I. Martini, pp , Kluwer Acad., Norwell, Mass. Paterson, M. (1978), Experimental Rock Deformation The Brittle Field, Springer, Berlin. Pauselli, C., and C. Federico (2002), The brittle/ductile transition along the Crop03 seismic profile: Relationship with the geological features, Boll. Soc. Geol. Ital., Vol. Spec. 1, Pauselli, C., R. Marchesi, and M. Barchi (2002), Seismic image of the compressional and extensional structures in the Gubbio area (Umbrian- Pre Apennines), Boll. Soc. Geol. Ital., Vol. Spec. 1, Pondrelli, S., S. Salimbeni, G. Ekstroem, A. Morelli, P. Gasperini, and G. Vannucci (2006), The Italian CMT dataset from 1977 to the present, Phys. Earth Planet. Inter., 159(3 4), Ponziani, F., R. De Franco, G. Minelli, G. Biella, C. Federico, and G. Pialli (1995), Crustal shortening and duplication of the Moho in the northern Apennines: A view from seismic refraction data, Tectonophysics, 252, Regione Umbria (2002), Rilevamento Geologico e Geotematico delle Aree Terremotate, Perugia, Italy. Sage, L., A. Mosconi, I. Moretti, E. Riva, and F. Roure (1991), Cross sections balancing in the central Apennines: An application of LOCACE, Am. Assoc. Pet. Geol. Bull., 75, Scarsella, F. (1951), Un aggruppamento di pieghe dell Appennino Umbro-Marchigiano, Boll. Serv. Geol. Ital., 73, Scholz, C. (1990), The Mechanics of Earthquakes and Faulting, Cambridge Univ. Press, Cambridge, U.K. Sibson, R. (1984), Roughness at the base of seismogenic zone: Contributing factors, J. Geophys. Res., 89(B7), Stucchi, E., F. Mirabella, and M. Ciaccio (2006), Comparison between reprocessed seismic profiles seismologic and geologic data: A case study of the Colfiorito earthquake area, Geophysics, 71, B29 B40. Waldhauser, F., and W. Ellsworth (2000), A double-difference earthquake location algorithm: Method and application to the northern Hayward fault, California, Bull. Geol. Soc. Am., 90, of 15