ClickHere

for

FullArticle

Normal faults and thrusts reactivated by deep fluids:The 6 April 2009 Mw 6.3 L’Aquila earthquake, central Italy

F. Di Luccio,1 G. Ventura,1 R. Di Giovambattista,2 A. Piscini,2 and F. R. Cinti1

Received 7 December 2009; revised 28 January 2010; accepted 17 February 2010; published 24 June 2010.

[1] On 6 April 2009, a Mw = 6.3 earthquake occurred in the central Apennines (Italy)damaging the city of L’Aquila and the surrounding country. We relocate the October 2008to 6 April 2009 foreshocks and about 2000 aftershocks occurred between 6 and 30 April2009 by applying a double‐difference technique and determine the stress field fromfocal mechanisms. The events concentrate in the upper 15 km of the crust. Three mainNW‐SE to NNW‐SSE striking, 30°–45° and 80°–90° dipping faults were activatedduring the seismic sequence. Among these, a normal fault and a thrust were reactivatedwith dip‐slip movements in response to NE‐SW extension. The structural maturity ofthe seismogenic fault system is lower than that displayed by other systems in southernApennines because of the lower strain rate of the central sector of the chain with respect tothe southern one. VP/VS increases progressively from October 2008 to the 6 April 2009main shock occurrence along a NW‐SE strike because of an increment in pore fluidpressure along the fault planes. Pore pressure diffusion controls the space‐time evolutionof aftershocks. A hydraulic diffusivity of 80 m2 s−1 and a seismogenic permeability ofabout 10−12 m2 suggest the involvement of gas‐rich (CO2) fluids within a highly fracturedmedium. Suprahydrostatic, high fluid pressure (about 200 MPa at 10 km of depth)within overpressurized traps, bounded by preexisting structural and/or lithologicaldiscontinuities at the lower upper crust boundary, are required to activate the April 2009sequence. Traps are the storage zone of CO2‐rich fluids uprising from the underlying,about 20 km deep, metasomatized mantle wedge. These traps easily occur in extensionalregimes like in the axial sector of Apennines but are difficult to form in strike‐slipregimes, where subvertical faults may cross the entire crust. In the Apennines, fluidsmay activate faults responsible for earthquakes up to Mw = 5–6. Deep fluids more thantectonic stress may control the seismotectogenesis of accretionary wedges.

Citation: Di Luccio, F., G. Ventura, R. Di Giovambattista, A. Piscini, and F. R. Cinti (2010), Normal faults and thrustsreactivated by deep fluids: The 6 April 2009 Mw 6.3 L’Aquila earthquake, central Italy, J. Geophys. Res., 115, B06315,doi:10.1029/2009JB007190.

1. Introduction

[2] On 6 April 2009, a destructive (about 300 casualties),Mw = 6.3 earthquake hits the city of L’Aquila and the centralsector of the Apennines, Italy (Figures 1a and 1b). Addi-tional damage was produced by the two larger aftershockslocated to the south (7 April, Mw = 5.6) and north (9 April,Mw = 5.4) of the epicentral area. The Neogene Apenninesfold‐and‐thrust belt represents the accretionary wedge of asubduction zone that includes the Tyrrhenian back‐arc basinto the west and the Adriatic‐Apulian foreland and foredeepto the east (Figure 1a) [Malinverno and Ryan, 1986;Doglioni, 1991]. At present, the central sector of the chain is

affected by a NE‐SW striking extension and uplift (up to2.5–3.0 mm yr−1) [Hunstad et al., 2003]. This extension isresponsible for the formation of intramountain basinsbounded by NW‐SE striking normal faults [Patacca et al.,2008]. The seismic activity of Apennines concentrates inthe axial sector of the chain (http://emidius.mi.ingv.it/CPTI08/) (Figure 1c) [Ventura et al., 2007]. The geody-namic significance of such seismicity is still debated anddifferent, not necessarily in conflict, hypotheses have beendone: (1) following Chiarabba et al. [2005, and referencestherein], the Apennines earthquakes are related to thenortheastward retreat of the Adria‐Ionian lithosphere; (2)according to Patacca et al. [2008], the seismic activity is dueto the gravity adjustment of the upper crust related to anincrease of the structural relief caused by an out‐of‐sequencepropagation of active thrusts at depth; (3) Lavecchia et al.[2003] propose that the Apennines earthquakes reflectrifting processes associated to large‐scale plume‐inducedlithospheric stretching in the Tyrrhenian domain; and/or

1Istituto Nazionale di Geofisica e Vulcanologia, Sismologia eTettonofisica, Rome, Italy.

2Istituto Nazionale di Geofisica e Vulcanologia, Centro NazionaleTerremoti, Rome, Italy.

Copyright 2010 by the American Geophysical Union.0148‐0227/10/2009JB007190

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 115, B06315, doi:10.1029/2009JB007190, 2010

B06315 1 of 15

Figure 1. (a) Structural scheme of Italy and seismicity distribution (red dots) from 1981 to 2002[Castello et al., 2005]. (b) Geological sketch map of the Abruzzi region (modified from Satolli andCalamita [2008]). CO2 gas fluxes are from Chiodini et al. [2000]. The low‐Vp zone at depth >20 kmis from Di Stefano et al. [1999]. (c) Crustal profile from CROP 11 data (modified from Ghisetti andVezzani [2002]).

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

2 of 15

(4) following other authors, the seismicity is due to theupward and eastward migration in the crust of CO2‐richfluids from a partly metasomatized mantle wedge beneaththe chain axis [Ghisetti and Vezzani, 2002; Chiodini etal., 2004; Ventura et al., 2007; Frezzotti et al., 2009].The role of fluids in northern Apennines is emphasized bya recent study by Miller et al. [2004], which proposes thataftershocks of the 1997 crustal earthquakes were drivenby the coseismic release of fluids through ruptures createdby the larger events.[3] Here, we study the Mw

max = 6.3 April 2009 L’Aquilaseismic sequence in the central Apennines. Previous studies[Chiarabba et al., 2009] analyze the distribution of the 712better localized events and conclude that a poorly knownnormal fault accommodates the extension of the area. Welocate the foreshocks and about 2000 aftershocks using adouble difference method, analyze the spatial distribution ofthe events, determine the stress field, and study the VP/VS

variations. The collected data and results are discussed inlight of the available geological and geophysical informa-tion and allow us to put constraints on (1) the activation ofinherited faults (e.g., preexisting thrusts), (2) the role of deepfluids in the nucleation process, and (3) the active geody-namics in accretionary wedges.

2. Geological Setting and Seismotectonics

[4] The Apennines chain resulted from the contempora-neous opening of the Tyrrhenian Sea, the eastward migra-tion of a compressive front, and the retreat of thelithospheric plate dipping below the Italian peninsula(Figure 1a) [Malinverno and Ryan, 1986; Doglioni, 1991,1995]. Due to the eastward migration of the compressivefront since the early Miocene, back‐arc extension affectedareas which were previously controlled by compressivetectonics. Evidence of these compressive tectonics is re-presented by NW‐SE striking thrusts, which place the car-bonate Meso‐Cenozoic succession in contact with theMiocene arenaceous flysch. The subsequent extensionaltectonics has been conditioned in many cases by thegeometry of the older thrust systems, with reactivation ofpreexisting structures [Galadini and Galli, 2000]. Since thePliocene, NW‐SE striking normal faults have been respon-sible for the formation of large intermountain basins inwhich Plio‐Quaternary continental sediments deposited.While compressive structures (overthrusts) characterize theApennines front, normal faults affect the Apennines chainfrom Pleistocene time [Doglioni, 1995]. Data from a NE‐SW striking seismic profile located 35 km south of the 2009L’Aquila seismic sequence evidence nappes of Mesozoic‐Triassic carbonates displaced by low‐angle thrusts that alsoinvolve lower Pliocene terrains (Figures 1b and 1c) [Ghisettiand Vezzani, 2002; Patacca et al., 2008; Di Luzio et al.,2009]. The upper Pliocene‐Quaternary deposits and theunderlying units are cut by normal faults, extending in depthto 10–15 km, with dip ≥ 45°. A Moho doubling under thecentral Apennines reflects the geometry of the mantle wedgebetween the subducting Adriatic lithosphere and the Apen-nines chain (Figure 1c). The April 2009 L’Aquila seismicsequence occurred in the central sector of Apennines, in anarea characterized by anomalous low compressional velocity(VP) and low attenuation (Q) at depth larger than 20 km

probably due to a fluid‐rich zone and heating from theunderlying mantle wedge (Figure 1b) [Mele et al., 1996; DiStefano et al., 1999]. According to Chiodini et al. [2000],the aquifers of this area are affected by a relevant, mantle‐derived CO2 flux with values in the order of 106 mol km−2

yr−1 (Figure 1b).[5] L’Aquila basin is bounded by the Gran Sasso and

Monte d’Ocre ranges, and by normal faults delimiting in-tramountain subbasins (Figure 2). The main geological unitsof the area can be summarized as follows. Jurassic‐Miocenelimestones and marls, and Miocene sandstones represent thebedrock outcropping on the ridges and valley flanks. Qua-ternary deposits include Pleistocene breccias and lacustrinedeposits [Blumetti et al., 2002]. Local debris alluvial fansoccur at the foot of the valley.[6] On the eastern side of the Aterno river valley (Figure 2),

southwestern dipping faults occur while antithetic faultsaffect the western side. The active faults strike (N)NW‐(S)SE, and dip 45° to 70° [Galadini and Galli, 2000] (Figure 2).They are characterized by normal to normal‐oblique slips andmove in response to a NE‐SW extension, which is acting atregional scale [Montone et al., 2004, and references therein].Preexisting, NNE‐SSW and ESE‐WNW to NW‐SE, low‐angle (dip < 45°) structures also outcrop (Figure 1b) [Pizziand Galadini, 2009]. The structural arrangement of theL’Aquila area results from the superimposition of the Qua-ternary extensional tectonics (early Pleistocene to date) to theNeogene compressive one [Ghisetti and Vezzani, 1999].[7] The early Quaternary surface faulting associated to

earthquakes larger than M > 5.5 mostly occurred on theNW‐SE striking faults. It is noteworthy that one of the NW‐SE normal faults bordering the northern side of the Aternobasin, i.e., the Paganica fault, reactivated during the April2009 L’Aquila sequence and surface faulting was observedfor a length of about 3 km [EMERGEO Working Group,2010].[8] Historical seismicity (http://emidius.mi.ingv.it/CPTI08/)

indicates that the region was affected by destructive earth-quakes (Figure 2). The main events occurred in 1349 AD(Me = 6.5), 1461 (Me = 6.4), 1703 (Me = 6.7), and 1762(Me = 6.0). Tertulliani et al. [2009] observe strong analogiesin the intensity distribution and the areas of the strongesteffects produced by the historical shocks and those occurredon April 2009. Bagh et al. [2007] report that the backgroundseismicity of the area is characterized by earthquakes withML ≤ 3.7 and it is locally clustered. The seismogenic volumeaffects the upper 15 km of the crust.

3. Seismic Sequence and Event Location

3.1. Methods

[9] About 20,000 aftershocks were recorded by thenational seismic network of the Istituto Nazionale di Geo-fisica e Vulcanologia (INGV) up to September 2009. Weused the P and S wave readings from the INGV bulletin andrelocated the seismicity occurred within one month from themain shock (6 April 2009) and the foreshocks from October2008 to 6 April 2009. We used HYPOINVERSE code[Klein, 2000] to get a first located data set and then appliedthe double‐difference technique (HypoDD by Waldhauserand Ellsworth [2000]) to better constrain the seismicity.The HypoDD method bases on a two‐step process: in the

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

3 of 15

first step, the traveltime differences for event pairs at com-mon stations are derived from the analysis of the catalogdata; then in the second step the computed differential tra-veltime data are used to calculate double‐difference hypo-central locations. HypoDD technique considers only eventswith a number of observations greater than a fixed value,and they are grouped into clusters of well‐linked earth-quakes (events belong to only one cluster and that are notconnected to other clusters) to insure stability of the inver-sion. A 1‐D velocity model from Bagh et al. [2007] in theinitial location as well as in the relocation procedure is used,with a VP/VS of 1.86. In this study, the minimum number ofobservations per event pair is 8, and the maximum hypo-central separation allowed between linked events was fixedat 10 km. We obtain traveltime differences for each eventpair with a separation distance <10 km at stations locatedwithin 60 km distance from the cluster centroid. We use Pand S wave picks equally weighted from 32 seismic stationsof the INGV seismic network, and for most of the after-shocks, readings from temporary networks installed after themain shock occurrence are also included. The INGV loca-tion of the 6 April 2009 main shock is 42.35°N, 13.38°E,and focal depth 9.46 km (http://portale.ingv.it).

[10] We relocate more than 2000 events. Seventy‐fivepercent of the double‐difference locations have uncertainty<300 m in horizontal direction, and 400 m is the depthresolution (Dz); for the remaining 25%, the horizontal res-olution is <400 m, while Dz is <600 m.

3.2. Results

[11] The obtained relocations, shown in Figure 3, aregrouped in a narrow zone both in depth and map view.Figure 3a shows three main clusters of seismicity. Theepicenter area evidences a roughly NW‐SE aligned seis-micity with focal depths mostly between 5 and 12 km(Figure 3b, profile 3). The northernmost cluster, which in-cludes theMw = 5.4 aftershock of 9 April, is NW‐SE oriented.South of the 6 April main shock, a dense area of seismicity,including the Mw = 5.6 aftershock of 7 April is roughlyNNW‐SSE oriented. The focal depths are generally deeperthan 6 km (Figure 3b, profile 4). It is noteworthy that the threerecognized clusters are spatially well localized, and rareseismicity characterizes the intercluster areas. Foreshocksconcentrate immediately south of the 6 April main shock andtheir focal depth extend from 4 to 12 km (Figure 3b).

Figure 2. Structural map of the L’Aquila area with evidenced Quaternary faults (modified fromEMERGEO Working Group [2010]). The historical earthquakes are from http://emidius.mi.ingv.it/CPTI08/.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

4 of 15

Figure 3

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

5 of 15

[12] Profiles 1, 2, and 3 in Figure 3b are NE‐SW oriented,while profile 4 strikes WSW‐ESE. In profile 1, which has awidth of 10 km, the seismicity depicts three subverticalclusters with depths reaching 12 km with the strongeraftershock at 11 km. The westernmost cluster includes theevents belonging to the northwestern tip of the 6 April mainshock area, whereas the other two clusters include eventsassociated to the 9 April Mw = 5.4 aftershock. Sparseearthquakes with depth between 12 and 18 km also occur.[13] Profile 2 (Figure 3b) has a width of 6 km. Two dense

subvertical clusters are identified: the westernmost one in-cludes most of the 6 April aftershocks with depths roughlybetween 7 and 11 km, whereas the easternmost one is slightlyshallower. Between the two clusters, earthquake distributionreveals two planes dipping about 45° and 30° toward SW.[14] The earthquakes on profile 3 (width 10 km) in

Figure 3b depict a 45°, southwest dipping plane between 11and 7 km of depth. The prolongation of this plane, whichincludes the 6 April main shock, to the surface intercepts thetrace of the Paganica fault, where alignment of continuoussurface breaks is observed [EMERGEO Working Group,2010]. From 7 to 5 km of depth, the dip of the abovedescribed plane decreases to about 20°–30°. Finally, aminor group of few events is depicted at about 10 km ofdepth, 5 km away from the main shock. The 6 April fore-shocks (Figure 3b) are between 9 and 11 km of depth, withfew shallower events describing a vertical cluster and otherfew events located northeastward of the 6 April main shockat about 7 km of depth.[15] Profile 4 in Figure 3b has a width of 4 km, and it is

centered on the 7 April Mw = 5.6 aftershock. The seismicityconcentrates between 7 and 12 km in depth, with the 7 Aprilaftershock at 14 km. As evidenced in the profile 4 (Figure 3b),the seismicity depicts a dense cloud of events without pre-ferred alignments.[16] As a consequence, the rupture plane associated to the

7 April event could be either the approximately E‐W ori-ented fault plane or the N‐NW oriented plane (see focalmechanism in Figure 3a). A recent study by Pino and DiLuccio [2009] reveals, however, that the preferred ruptureplane is the N‐NW oriented, which is also consistent withthe epicentral distribution of aftershocks (Figure 3a).

4. Stress Field

[17] Focal mechanisms of the larger foreshocks andaftershocks of the April 2009 sequence (Mw > 3), availableon line at http://eqinfo.eas.slu.edu/Earthquake_Center/MECH.IT/, were analyzed. The focal solutions of theaftershocks are consistent with normal slips (Figure 3a). Thesolutions of the foreshocks indicate normal and obliqueslips. Nodal planes of 35 focal mechanisms show a clearNW‐SE preferred strike along planes dipping 45° toward SE

(Figures 4a and 4b). A minor N‐S strike is also present. Thepreferred strike is consistent with the elongation of the6 April 2009 aftershock zone (Figure 3a). The N‐S strike isroughly compatible with the elongation of the aftershocksassociated to the 7 April Mw = 5.6 event. We use 35 focalmechanisms of significant earthquakes to compute the stressfield, by applying the method developed by Michael [1984].Results are reported in Figure 4c and indicate a normal stressfield characterized by a subhorizontal, NE‐SW strikingminimum compressive stress s3. This stress configuration isconsistent with that acting in the central Apennines [Montoneet al., 2004; Bagh et al., 2007].

5. VP /VS

[18] Foreshocks of the 6 April main shock started mainlyfrom October 2008 and concentrated at about 9–11 km ofdepth (Figure 3b). The cumulative number of events as afunction of time prior to the main shock clearly shows asignificant increase in the total number of recorded earth-quakes (Figure 5). Three main trends of seismicity ratedefined by jumps in the cumulative number of events areevident. We estimated VP/VS from the foreshocks by apply-ing the modified Wadati method [Wadati, 1933] to each ofthe identified three trends with constant seismic rate. Thestarting data set including P and S phases with weightranging between 0 and 4 was restricted to a subset of clear,sharp onsets most of which having 0 weight. Taking intoaccount the real uncertainty on DTP and DTS, the obtainedVP/VS was fitted using a linear least squares regression with acorrelation coefficient varying between 0.92 and 0.98 for thetime periods in Figure 5 and between 0.78 and 0.94 for theAQU‐FIAM and AQU‐FAGN paths in Figure 6. The lowestvalue of 0.78 was obtained for the period March 2009 to6 April 2009 relative to the AQU‐FAGN path (Figure 6),whose data are more sparsely distributed with respect tothe other path and/or time periods. We selected the arrivaltimes from the nearest stations distributed around theepicentral area. The Wadati plot (Figure 5) referring to theOctober 2008 to January 2009 period shows a VP/VS =1.83 ± 0.01. In the period January 2009 to March 2009,VP/VS is 1.84 ± 0.01. This last value slightly increases to1.86 ± 0.01 in the time interval 31 March to 6 April2009. From the main shock occurrence to 30 April, VP/VS

reaches a value of 1.95 ± 0.01 (Figure 5). In conclusion,VP/VS gently increases from October 2008 to 6 April2009, and abruptly changes from the main shock (6 April,0132 LT) to 30 April (Figure 5).[19] In order to study the VP/VS changes along different

strikes, we selected data relative to two different structuralpaths (Figure 6): (1) a NW‐SE orientation, which coincideswith the preferred strike of the Aterno faults and aftershockalignment (AQU‐FAGN), and (2) a NE‐SW orientation

Figure 3. (a) Time and epicentral distribution of the April 2009 L’Aquila seismic sequence and its foreshocks. Stars indicatethe foreshocks occurred from 1 October 2008 to 6 April 2009 (0120 LT). Focal mechanisms of the events withMw > 5 (blackbeach balls) and of the foreshocks (gray beach balls) are from http://eqinfo.eas.slu.edu/Earthquake_Center/MECH.IT/. Faultsare from Figure 2. (b) Cross sections of the seismicity depicted in Figure 3a, with black dots indicating the aftershocks and thered stars indicating the foreshocks. The yellow circle and star and the green diamond correspond to the events marked inFigure 3a. Quaternary faults (red segments, dashed when the dip is inferred) are from Figure 3a, while the trace of the GranSasso thrust (blue segments in profiles 2 and 3) is from Satolli and Calamita [2008]. In profile 3, the crustal layers are fromFigure 1c.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

6 of 15

(AQU‐FIAM) roughly transversal to the previous one.There is a large increase of VP/VS estimated along the pathAQU‐FAGN from the January 2009 to March 2009 periodto the following March to 6 April 2009 (Figure 6). In par-ticular, VP/VS increases from 1.87 ± 0.01 to 1.97 ± 0.02 forthe AQU‐FAGN path. To the contrary, VP/VS on the AQU‐FIAM path is nearly constant with values of 1.77–1.78.

6. Discussion

6.1. Activated Faults and Kinematics

[20] The distribution of the April 2009 L’Aquila seismicsequence (Figure 3) indicates that earthquakes occurredalong three main structures.

[21] Structure A is a 15 km long, NW‐SE striking, about50° SW dipping structure (6 April main shock and after-shocks) from which a 30° SW dipping structure departs atabout 7–8 km of depth (profiles 2 and 3 in Figure 3b). Theseismicity mainly occurs at the boundary between an up-thrusted block of the metamorphic basement and the over-lying nappes and folds of the chain, in an area where theuprising of the mantle wedge occurs (profile 3 in Figure 3b)[Ghisetti and Vezzani, 2002]. The location of the aftershocksoverlaps the Paganica fault trace (Pf in Figure 3b), as alsofound in other studies [Chiarabba et al., 2009]. The surfaceruptures associated to the 6 April main shock along thePaganica fault [EMERGEO Working Group, 2010] stronglysupports this interpretation. Our data show, however, that

Figure 4. (a) Strike, dip, and rake distribution of focal mechanism nodal planes (data from http://eqinfo.eas.slu.edu/Earthquake_Center/MECH.IT/). (b) Density distribution of poles to nodal planes. (c) Resultsof the stress field analysis [Michael, 1984] on 35 focal mechanisms of earthquakes with Mw > 3 occurredin April 2009. The parameter � = (s2 − s3)/(s1 − s3), with the principal stress axes s1 ≥ s2 ≥ s3.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

7 of 15

Figure 5. Time versus cumulative number of earthquakes for the period October 2008 to April 2009,including the main shock of 6 April 2009 (yellow star). VP/VS calculated for the different time intervalsmarked by jumps in the cumulative number of events are reported with different colors. VP/VS values weredetermined using the stations closest to the 6 April main shock location. The linear best fit is reported ascontinuous line, the dashed lines are the 95% prediction limit.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

8 of 15

Figure 6. VP/VS values calculated for two different time intervals along the AQU‐FIAM and AQU‐FAGNpaths, which are shown in the map along with faults (red segments) from Figure 2.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

9 of 15

the Gran Sasso thrust was likely reactivated (GSt inFigure 3b). The rake distribution from the focal mechanisms(Figure 4) indicates prevailing normal slips along NW‐SEstriking, with 30° to 50° SW dipping rupture planes.Therefore, both the Gran Sasso thrust, whose activity datesback to pre‐Pliocene times, and the Quaternary Paganicafault were reactivated as normal faults during the April 2009sequence. The foreshock distribution indicates that the earlystage of the rupture process occurred on the Paganica fault,and in particular, on its hanging wall. This observation wellagrees with numerical models on the early stages of co-seismic fault activation [Zhang and Sanderson, 1996a]. Suchmodels show that, within a prefractured medium crossed by anormal fault, the rupture affects both the fault zone and thehanging wall. As the April 2009 sequence evolves in time,aftershocks concentrate in the fault zone and in its footwall.[22] Structure B is a 20 km long, NNW‐SSE striking, 4–

5 km wide rupture zone, confined between 6 and 12 km ofdepth, is evidenced by the 7 April 2009 (Mw = 5.6) largeraftershock and associated seismicity (Figure 3a and profile4 in Figure 3b). On the surface, faults possibly related tothis rupture are lacking.[23] A gap in seismicity exists between the 7 April (Mw =

5.6) nucleation, which occurred at ∼14 km of depth, and therelated cluster (profile 4 in Figure 3b). This gap could bedue to an aseismic shearing zone. Aseismic shearing mayoccur in the middle‐lower continental crust, where it isgenerally associated to fluid trapped by impermeable layers[Goto et al., 2005]. In the L’Aquila area (Figure 1b), evi-dence of pressurized, mantle‐derived fluids is given by thehigh CO2 release (1–5 × 106 mol km−2 yr−1) [Chiodini et al.,2000], while the impermeable layer could be represented bythe Permo‐Triassic evaporites, occurring between themetamorphic basement of the chain and the overlyingnappes, at depth between 12 and 16 km [Patacca et al.,2008]. As a preliminary hypothesis, we suggest that aseis-mic shearing occurred in the seismic gap area between the7 April (Mw = 5.6) event and the overlying aftershockvolume.[24] Structure C consists of two subvertical, 15–20 km

long NW‐SE striking structures extending between 3 and4 km and about 12 km of depth are depicted by the 9 April2009 (Mw = 5.4) event and its aftershocks (profile 1 inFigure 3b). Evidence of surface faulting for these structuresis lacking [Atzori et al., 2009; EMERGEO Working Group,2010].[25] The low number of seismic events among the seis-

mogenic structures A, B, and C, whose activity was con-tinuous from 6 April to the end of April 2009, indicates thelack of mature transfer zones among the main NW‐SE toNNW‐SSE striking ruptures. According to structural modelsof normal fault networks [Peacock, 2002], this featuresuggests an early stage of formation for the L’Aquila seis-mogenetic fault network. On the basis of above data andobservations, we propose that the L’Aquila sector of theApennines is at a less mature stage with respect to thesouthern sector of the chain, where, instead, seismicity ontransfer structures among the main NW‐SE faults was rec-ognized [Milano et al., 2002]. The lower structural maturityof the central Apennines could be due to a relative lowerextension rate of the central sector of the chain, which is

roughly about 3 mm yr−1, whereas values up to 5 mm yr−1

characterize the southern Apennines [Hunstad et al., 2003].[26] The focal mechanisms of the L’Aquila events with

Mw > 5 (Figure 3), as well as the results of the stress fieldanalysis (Figure 4b) show that the recognized A, B, and Cseismogenetic structures moved in response to a NE‐SWextension, which is that acting at regional scale, as previ-ously reported [Montone et al., 2004]. Therefore, the faultkinematics and the stress field of the L’Aquila sequence areconsistent with the present‐day tectonic configuration of theApennines [Chiarabba et al., 2005; Mantovani et al., 2009].Patacca et al. [2008] suggest that the NE‐SW extension andApennines seismicity are related to the gravity adjustment ofthe upper crust due to an out‐of‐sequence propagation of theactive thrusts, i.e., compression, at depth. The focal me-chanisms of the L’Aquila 2009 events do not evidenceactive thrusting (compression), showing normal and minoroblique slip solutions (Figure 4). On the basis of our results,we exclude the hypothesis by Patacca et al. [2008] on theoccurrence of active compression within the axial sector ofcentral Apennines.

6.2. Evidence of Fluids and Preferred Fluid Path

[27] Some authors [Ghisetti and Vezzani, 2002; Chiodiniet al., 2004; Ventura et al., 2007] suggest that the seismo-tectogenesis of the Apennines is related to the uprising ofmantle derived fluids. In addition, Miller et al. [2004] pro-pose that the aftershocks of the Umbria‐Marche 1997sequence in northern Apennines were related to the releaseof overpressurized CO2.[28] The VP/VS value estimated for the foreshocks and

aftershocks of the L’Aquila 2009 sequence is between 1.83and 1.95 and increases from October 2008 to April 2009(Figure 5). VP/VS = 1.83 was also found in previous studieson central Apennines [Bagh et al., 2007], whereas VP/VS =1.95 is anomalously high, and it is consistent with valuesestimated in fluid‐rich zones [Zhao and Negishi, 1998;Husen and Kissling, 2001]. In particular, a VP/VS value of1.9 was found in northern Apennines and was interpreted toreflect the presence of pressurized fluids in the crust[Moretti et al., 2009]. We conclude that fluids were presentwithin the seismogenic volume of the L’Aquila 2009sequence.[29] Laboratory and borehole experiments proved VP/VS

to increase with the concentration of fluid‐saturated cracks[Moos and Zoback, 1983], and several studies showanomalous changes in VP/VS during earthquakes [Nadeau etal., 1994; Chen et al., 2001]. An increase in VP/VS mayevidence the presence of fluids in the seismogenic faults[Eberhart‐Phillips and Michael, 1993; Zhao et al., 1996].Therefore, the raise in seismicity rate and VP/VS during theL’Aquila 2009 sequence (Figure 5) could be due to anincrease of cracking associated to fluid migration. Theoccurrence of structurally controlled, i.e., NW‐SE to NNW‐SSE striking, subvertical clouds of events (profiles 1, 3, and4 in Figure 3b) is also compatible with a migration of fluids,being these vertical clouds typical of fluid‐induced seis-micity [Shapiro et al., 1997, 2003], as also recognized insequences in the northern Apennines associated to fluidmovements [Calderoni et al., 2009].[30] Data reported in Figure 6 show a significant increase

of VP/VS (from 1.87 to 1.97) along the AQU‐FAGN path,

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

10 of 15

which is parallel to the NW‐SE striking Paganica fault.Therefore, the seismogenic volume is characterized by atime‐ and space‐dependent anisotropy, which can reflect thepreferred migration of fluids along the strike of the rupturezone. This interpretation well agrees with numerical models[Zhang and Sanderson, 1996b] and field observations on themigration of fluids in active faults [Sibson, 2000, and re-ferences therein]. Such models predict that the permeabilityof faults increases along the direction of the maximum hor-izontal stress sH and fluid flow is allowed along this direc-tion. In an extensional regime, sH = s2 and fluids preferablymigrate along planes containing s2. In the L’Aquila case, s2is parallel to the preferred NW‐SE fault strike (Figure 4). Inconclusion, the increase of VP/VS along the AQU‐FAGNpath (Figure 6) results from the passage of fluids along theNW‐SE fault strike possibly allowed by the higher perme-ability of the faulted rocks. This interpretation is consistentwith results from previous studies, which show that lateralvariations of VP/VS are sensitive to the faulting property[Eberhart‐Phillips and Michael, 1998; Graeber and Asch,1999; Gentile et al., 2000]. As concerns the role of fluidsin activating preexisting structures, we remark that whilestatic stress interaction explains failure between collinearsegments of strike‐slip faults or collateral segments of nor-mal faults, it does not hold for the activation of collinearnormal faults [Nostro et al., 1997]. According to Noir et al.[1997] and Sibson [2000], the interaction and activation inshort times (few hours) of collinear normal faults like thoseactivated during the L’Aquila sequence may arise from thepropagation of a fluid pressure wave through an anisotropicfractured rock mass.

6.3. Pore Pressure Diffusion and Permeability

[31] Shapiro et al. [1997, 2003] propose that the trigger-ing of seismicity and the consequent spatiotemporal evolu-tion can be analyzed in terms of pore pressure relaxation inmedia with (an)isotropic hydraulic diffusivity. In a plotdistance of the pressure front from the triggering front ver-sus time, a parabolic‐like envelope of cloud of events isrecognized when a diffusion‐like process operates. In aporoelastic medium, the extension of the rupture zone canbe approximated, by the theoretical curve r = (4pDt)1/2,where the distance r of the pressure front from the fluidsource (triggering front) is as a function of the diffusivity Dand time t [Shapiro et al., 1997]. We analyzed (1) theforeshock distribution from 30 March 2009 to 6 April 2009before the main shock because of a marked increase in thecumulative number of events (Figure 5) and (2) the after-shocks from the main shock to 30 April 2009. Both theforeshocks and the aftershocks used in this analysis are therelocated events described in profile 3. In a distance‐versus‐time (r‐t) diagram, we identify a triggering front with adiffusivity of 4.5 m2 s−1 for the foreshocks (Figure 7a),whereas the aftershocks give a diffusivity of 80 m2 s−1

(Figure 7b). We used a linearized inversion procedure to fitthe equation r = (4pDt)1/2 to hypocenter data followingSaccorotti et al. [2002]. In particular, our procedure forfitting the above equation is to select the farthest earth-quakes that occurred in consecutive, not overlapping times.We remark that, while a clear diffusion‐like seismicity isdepicted by the aftershocks, the low number of foreshocksdoes not allow us to unequivocally identify a fluid‐con-

trolled seismicity. In any case, the determined D values arewithin those reported in literature [e.g., Roeloffs, 1996;Talwani et al., 2007] with D = 80 m2 s−1 representing anupper limit [Roeloffs, 1996]. This latter, large value of Dimplies a high‐permeability k of the crustal rocks, being Dproportional to k [e.g., Talwani et al., 2007] through therelation:

k ¼ D�F� ð1Þ

where h, F, and b are the fluid viscosity, the porosity of therock, and the compressibility of fluid, respectively. In thecase of water, h = 10−3 Pa s, b = 3 × 10−9 Pa−1. F is set to0.07 [Iscan et al., 2006]. In presence of CO2, the product hbdoes not change because CO2 at source depth is in asupercritical state, and it is 10 times more compressible thanwater and 10 time less viscous [Miller et al., 2004]. Usingthe above selected parameters and equation (1), we obtaink = 5.6 × 10−12 m2, which has the significance of an orderof magnitude estimate and refers to a dynamic, seismogenicpermeability [Talwani et al., 2007]. This value is 2 orders ofmagnitude larger than that measured in fault gauges relativeto faults outcropping in the Abruzzo area (k ≤ 10−14 m2)[Agosta et al., 2007]. Besides, the estimated value of per-meability is consistent with the gas permeability valuesmeasured in carbonatic rocks (k up to 10−12 m2) [Lucia,1999] and in laboratory experiments on limestones understress conditions (k up to 1.3 × 10−12 m2) [Iscan et al.,2006].

6.4. Fluid Pressure and Stress Regime and ComparisonWith Other Fluid‐Rich Structural Settings

[32] The above summarized results indicate that fluidpressure played an important role in the April 2009 seis-micity. In order to determine the fluid pressure necessary toactivate the A, B, and C L’Aquila structures (section 6.1),we calculate the stress difference s1‐s3 required to fracturethe rock along a plane with ideal orientation and comparethis value with that required for slip along different pre-existing planes of weakness of variable orientation, fol-lowing the approach of Yin and Ranalli [1992] asimplemented in the ReActiva software [Alaniz‐Alvares andTolson, 2000]. The input data include the stress ratio F therock cohesion c, the coefficient of internal friction m, and thepore fluid factor l = Pf /(rgz), with Pf fluid pressure, r rockdensity, g gravity, and z depth. In a normal stress regime,rgz = s1. For the April 2009 L’Aquila sequence, we adopt anormal stress regime with a NE‐SW striking s3 (Figure 4b),and set: F = 0.65 (see Figure 4), c = 0 assuming that theruptures occur along preexisting planes of weakness (for theintact rock, c = 10 MPa), m = 0.6 [Sibson, 2000], r = 2650 kgm−3, and z = 10 km (Figure 3). Results are shown in Figure 8,where the contour plot (gray areas) of the poles to planes isreported at different l as a function of the ratio T between theshear and normal stress [Morris et al., 1996]. In Figure 8,T/TS gives the slip tendency of each plane being TS themaximum calculated T. The poles to the inferred activatedstructures A (NW‐SE strike; dip 45°, Paganica fault and30°, SW dipping, reactivated Gran Sasso thrust), B (NW‐SE strike, dip 80°–90°), and C (NNW‐SSE strike, dip 80°–90°) are also reported. Results show that for l = 0.4, noneof the A, B, and C poles falls in the field of the slipping

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

11 of 15

Figure 7. (a) A r‐t plot for the 30 March 2009 to 6 April 2009 (0120 LT) foreshock relocated events.(b) A r‐t plot for the 6 to 30 April 2009 aftershocks relocated events. The spatiotemporal seismicitypattern shows vertically clustered events interrupted by time intervals of seismic quiescence or loweringin the seismic rate. The minimum ML is 1.0. In both Figures 7a and 7b the symbols are scaled withmagnitude ML.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

12 of 15

faults. For l = 0.6, the NW‐SE and NNW‐SSE strikingstructures with dip ≥ 45° fall in the contour plot with T/TS≥ 0.8, whereas the structure with dip = 30° is in the fieldwith T/TS ≤0.1. For l = 0.8, all the inferred rupture planesfall in the area with T/TS ≥ 0.8. These results suggest asuprahydrostatic fluid pressure. We conclude that l valuesbetween 0.6 and 0.8 are necessary to explain the (re)acti-vation of the April 2009 L’Aquila structures. Taking intoaccount the relation between l and Pf, we estimate a Pf

between 155 and 207 MPa at 10 km depth. The largest Pf

value is consistent with that calculated to activate thrusts inthe outer sector of Apennines (Pf = 215 MPa at 10 kmdepth from Calderoni et al. [2009]), where a compressivestress regime exists. We propose that, despite the differentstress regime and fluid sources, i.e., methane and petro-leum in the outer sector [Calderoni et al., 2009] andmantle‐derived CO2 in the axial sector [Chiodini et al.,2004], the fluid pressure below these two sectors ofApennines chain is nearly constant. This implies theexistence of overpressurized reservoirs in which, after anearthquake, Pf possibly decreases due to the upwardmigration of fluids along the activated fault(s). The fluidinvolvement in fault systems characterized by a normalstress regime like that recognized in the L’Aquila regionmay differ from that in strike‐slip regimes, e.g., the BohaiBay Basin system (China). Here, mantle‐derived fluidsmigrate along subvertical faults that cross the basement,and, in particular, at the intersection of faults, where fluidpressure is low because of the continuous gas release[Zhang et al., 2008].

6.5. Geodynamic Implications

[33] The April 2009 L’Aquila seismicity evidences su-prahydrostatic overpressures approaching lithostatic valuesin a structurally controlled intramountain basin (Figures 2and 8). As noticed by Sibson [2000], the sustainability oflarge fluid pressures in extensional regimes is a still debatedquestion, however, our data show that, in central Apennines,relatively large fluid pressures are required to activate faultsthat may produce Mw = 5–6 earthquakes. According to theavailable geochemical [Chiodini et al., 2004; Frezzotti et al.,2009] and geodynamic [Ghisetti and Vezzani, 2002; Venturaet al., 2007] models of the Apennines, the source of fluids

may be the metasomatized mantle wedge, which releasescarbon dioxide in the overlying continental crust. Our datasuggest that pressurized fluid traps within the crust may beconfined by structural (preexisting thrusts and folds withinthe carbonates) and/or lithological (impermeable layers likethe Permo‐Triassic evaporites) discontinuities. This geody-namic picture may explain why the seismicity of theApennines concentrates in the upper plate and not in thedowngoing plate or in both the downgoing and overridingplates, as observed in the majority of subduction settings[Stern, 2002].

7. Conclusions

[34] The results of this study may be summarized in thefollowing main points.[35] 1. The April 2009 L’Aquila seismic activity devel-

oped along three main, quasi‐collinear structures strikingNW‐SE to NNW‐SSE that moved in response to a NE‐SWextension. One of these structures corresponds to an out-cropping normal fault (Paganica). A portion of the GranSasso thrust, at the junction with the fault responsible of theearthquake, was also reactivated with a normal mechanism.The structural maturity of this system of structures is lowerthan that displayed by similar systems in southern Apen-nines, which are affected by a larger strain rate.[36] 2. VP/VS increases progressively from October 2008

to the 6 April 2009 main shock along a NW‐SE preferredstrike. This increase is related to an increment of pore fluidpressure. The VP/VS spatial anisotropy is related to themovement of fluids along fault planes.[37] 3. Pore pressure diffusion is the main mechanism

controlling the space‐time distribution of aftershocks. Anincrease of fluid pressure due to the upward migration offluids induced an increase of pressure in the connected porespace, which includes pores and cracks. By increasing thepore pressure the effective normal stress and cohesiondecrease. This leads to sliding along preexisting subcriticalcracks and to the initiation of the rupture. Hydraulic diffu-sivity is about 80 m2 s−1, which represents an upper boundfor the diffusivity of crustal rocks and which probably re-flects the involvement of gas (CO2) from deep source. Theseismogenic permeability is in the order of 10−12 m2. Su-

Figure 8. Results of the fault slip analysis. Areas with 0 ≤ T ≤ TS (TS = 1) are the theoretically predictedpatterns of poles to reactivated fault planes for different values of s1. Poles to the L’Aquila 2009 reacti-vated structures A, B, and C are also reported as dots.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

13 of 15

prahydrostatic pressure conditions were required to activatethe L’Aquila seismic sequence with Pf values up to about200 MPa at 10 km of depth.[38] 4. Overpressurized traps along preexisting structural

and/or lithological discontinuities at the lower‐upper crustboundary are required to explain the calculated Pf. Suchtraps may represent the storage zone of CO2‐rich fluidsuprising from the underlying, metasomatizied mantlewedge. These traps, which are easy to form in an extensionalregimes like that acting in the L’Aquila area, are difficult todevelop in strike‐slip regimes, where subvertical faults maycross the entire crust.[39] 5. In the L’Aquila zone of central Apennines, fluids

may activate faults producing earthquakes up to Mw = 5–6.The April 2009 L’Aquila sequence occurred on the axialzone of the chain, i.e., in the overriding plate of the Apen-nines subduction system, and not in the downgoing plate, as,on the contrary, usually recognized in subduction zones.Such features suggest that deep fluids more than tectonicstress control the seismotectogenesis of accretionary wedges.

[40] Acknowledgments. Robert Herrmann is gratefully thanked forproviding public access to the online focal mechanisms of the L’Aquila2009 sequence. Discussions with G. Chiodini, R. Devoti, and F. P. Lucenteand comments by the Associate Editor, an anonymous reviewer, andZhengfu Guo helped to improve the manuscript. Some of the figures weredone using GMT [Wessel and Smith, 1991]. G.V. was supported by theIYPE project Creep.

ReferencesAgosta, F., M. Prasad, and A. Aydin (2007), Physical properties of carbon-ate fault rocks, Fucino basin (central Italy): Implications for fault seal inplatform carbonates, Geofluids, 7, 19–32, doi:10.1111/j.1468-8123.2006.00158.x.

Alaniz‐Alvares, S., and G. Tolson (2000), Assessing fault reactivation withthe ReActiva Program, J. Geosci. Educ., 48, 651–657.

Atzori, S., I. Hunstad, M. Chini, S. Salvi, C. Tolomei, C. Bignami,S. Stramondo, E. Trasatti, A. Antonioli, and E. Boschi (2009), Finite faultinversion of DInSAR coseismic displacement of the 2009 L’Aquila earth-quake (central Italy), Geophys. Res. Lett., 36, L15305, doi:10.1029/2009GL039293.

Bagh, S., L. Chiaraluce, P. De Gori, M. Moretti, A. Govoni, C. Chiarabba,P. Di Bartolomeo, and M. Romanelli (2007), Background seismicity inthe central Apennines of Italy: The Abruzzo region case study, Tectono-physics, 444, 80–92, doi:10.1016/j.tecto.2007.08.009.

Blumetti, A. M., M. Di Filippo, P. Zaffiro, P. Marsan, and B. Toro (2002),Seismic hazard of the city of L’Aquila (Abruzzo‐central Italy): New datafrom geological, morphotectonic and gravity prospecting analysis, StudiGeol. Camerti, 1, 7–18.

Calderoni, G., R. Di Giovambattista, P. Burrato, and G. Ventura (2009), Aseismic sequence from northern Apennines (Italy) provides new insighton the role of fluids in the active tectonics of accretionary wedges, EarthPlanet. Sci. Lett., 281, 99–109, doi:10.1016/j.epsl.2009.02.015.

Castello, B., G. Selvaggi, C. Chiarabba, and A. Amato (2005), Catalogodella sismicità Italiana—CSI 1.0, 1981–2002, Cent. Naz. Terremoti,Ist. Naz. di Geofis. e Vulcanol., Rome.

Chen, C. H., W. H. Wang, and T. L. Teng (2001), 3D velocity structurearound the source area of the 1999 Chi‐Chi, Taiwan, earthquake: Beforeand after the mainshock, Bull. Seismol. Soc. Am., 91, 1013–1027,doi:10.1785/0120000737.

Chiarabba, C., L. Jovane, and R. Di Stefano (2005), A new view of Italianseismicity using 20 years of instrumental recordings, Tectonophysics,395, 251–268, doi:10.1016/j.tecto.2004.09.013.

Chiarabba, C., et al. (2009), The 2009 L’Aquila (central Italy) MW 6.3earthquake: Main shock and aftershocks, Geophys. Res. Lett., 36,L18308, doi:10.1029/2009GL039627.

Chiodini, G., F. Frondini, C. Cardellini, F. Parello, and L. Peruzzi (2000),Rate of diffuse carbon dioxide Earth degassing estimated from carbonbalance of regional aquifers: The case of central Apennine, Italy, J. Geo-phys. Res., 105, 8423–8434, doi:10.1029/1999JB900355.

Chiodini, G., C. Cardellini, A. Amato, E. Boschi, S. Caliro, F. Frondini,and G. Ventura (2004), Carbon dioxide Earth degassing and seismogen-esis in central and southern Italy, Geophys. Res. Lett., 31, L07615,doi:10.1029/2004GL019480.

Di Luzio, E., G. Mele, M. M. Tiberti, G. P. Cavinato, and M. Parotto(2009), Moho deepening and shallow upper crustal delamination beneaththe central Apennines, Earth Planet. Sci. Lett., 280, 1–12, doi:10.1016/j.epsl.2008.09.018.

Di Stefano, R., C. Chiarabba, F. P. Lucente, and A. Amato (1999), Crustaland uppermost mantle structure in Italy from the inversion of P wavearrival times: Geodynamic implications, Geophys. J. Int., 139, 483–498,doi:10.1046/j.1365-246x.1999.00952.x.

Doglioni, C. (1991), A proposal of kinematic modelling for W‐dippingsubductions—Possible applications to the Tyrrhenian‐Apennines system,Terra Nova, 3, 423–434, doi:10.1111/j.1365-3121.1991.tb00172.x.

Doglioni, C. (1995), Geological remarks on the relationships betweenextension and convergent geodynamic settings, Tectonophysics, 252,253–267, doi:10.1016/0040-1951(95)00087-9.

Eberhart‐Phillips, D., and A. J. Michael (1993), Three‐dimensionalvelocity structure and seismicity in the Parkfield region, central California,J. Geophys. Res., 98, 15,737–15,758, doi:10.1029/93JB01029.

Eberhart‐Phillips, D., andA. J.Michael (1998), Seismotectonics of the LomaPrieta, California, region determined from three‐dimensional Vp, Vp/Vs,and seismicity, J. Geophys. Res., 103, 21,099–21,120, doi:10.1029/98JB01984.

EMERGEOWorking Group (2010), Evidence for surface rupture associatedwith the Mw 6.3 L’Aquila earthquake sequence of April 2009 (centralItaly), Terra Nova, 22, 43–51, doi:10.1111/j.1365-3121.2009.00915.x.

Frezzotti, M. L., A. Peccerillo, and G. Panza (2009), Carbonate metasoma-tism and CO2 lithosphere‐asthenosphere degassing beneath the westernMediterranean: An integrated model arising from petrological and geo-physical data, Chem. Geol., 262, 108–120, doi:10.1016/j.chemgeo.2009.02.015.

Galadini, F., and P. Galli (2000), Active tectonics in the central Apennines(Italy)—Input data for seismic hazard assessment, Nat. Hazards, 22,225–268, doi:10.1023/A:1008149531980.

Gentile, G. F., G. Bressan, L. Burlini, and R. De Franco (2000), Three‐dimensional Vp and Vp/Vs models of the upper crust in the Friuli area(northeastern Italy), Geophys. J. Int., 141, 457–478, doi:10.1046/j.1365-246x.2000.00095.x.

Ghisetti, F., and L. Vezzani (1999), Depth and modes of Pliocene‐Pleisto-cene crustal extension of the Apennines (Italy), Terra Nova, 11, 67–72,doi:10.1046/j.1365-3121.1999.00227.x.

Ghisetti, F., and L. Vezzani (2002), Normal faulting, transcrustal perme-ability and seismogenesis in the Apennines (Italy), Tectonophysics,348, 155–168, doi:10.1016/S0040-1951(01)00254-2.

Goto, T., Y. Wada, N. Oshiman, and N. Sumitomo (2005), Resistivitystructure of a seismic gap along the Atotsugawa fault, Japan, Phys. EarthPlanet. Inter., 148, 55–72, doi:10.1016/j.pepi.2004.08.007.

Graeber, F. M., and G. Asch (1999), Three‐dimensional models of P wavevelocity and P‐to‐S velocity ratio in the southern central Andes by simul-taneous inversion of local earthquake data, J. Geophys. Res., 104,20,237–20,256, doi:10.1029/1999JB900037.

Hunstad, I., G. Selvaggi, N. D’Agostino, P. England, P. Clarke, andM. Pierozzi (2003), Geodetic strain in peninsular Italy between 1875and 2001, Geophys. Res. Lett., 30(4), 1181, doi:10.1029/2002GL016447.

Husen, S., and E. Kissling (2001), Postseismic fluid flow after the largesubduction earthquake of Antofagasta, Chile, Geology, 29, 847–850,doi:10.1130/0091-7613(2001)029<0847:PFFATL>2.0.CO;2.

Iscan, A. G., M. V. Kök, and A. S. Bagcı (2006), Estimation of permeabil-ity and rock mechanical properties of limestone reservoir rocks understress conditions by strain gauge, J. Petrol. Sci. Eng., 53, 13–24,doi:10.1016/j.petrol.2006.01.008.

Klein, F. (2000), User’s guide to HYPOINVERSE‐2000, a FORTRANprogram to solve for earthquake locations and magnitudes, U.S. Geol.Surv. Open File Rep., 02‐171, 123 pp.

Lavecchia, G., P. Boncio, and N. Creati (2003), A lithospheric scale seis-mogenic thrust in central Italy, J. Geodyn., 36, 79–94, doi:10.1016/S0264-3707(03)00040-1.

Lucia, J. (1999), Carbonate Reservoir Characterization: An IntegratedApproach, 2nd ed., 336 pp., Springer, Berlin.

Malinverno, A., and W. B. F. Ryan (1986), Extension in the TyrrhenianSea and shortening in the Apennines as result of arc migration drivenby sinking of the lithosphere, Tectonics, 5, 227–245, doi:10.1029/TC005i002p00227.

Mantovani, E., D. Babbucci, C. Tamburelli, and M. Viti (2009), A reviewon the driving mechanism of the Tyrrhenian‐Apennines system: Implica-tions for the present seismotectonic setting in the central‐northern Apen-nines, Tectonophysics, 476, 22–40, doi:10.1016/j.tecto.2008.10.032.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

14 of 15

Mele, G., A. Rovelli, D. Seber, and M. Barazangi (1996), Lateral variationsof Pn propagation in Italy: Evidence for a high attenuation zone beneaththe Apennines, Geophys. Res. Lett., 23, 709–712, doi:10.1029/96GL00480.

Michael, A. J. (1984), Determination of stress from slip data: Faultsand folds, J. Geophys. Res., 89, 11,517–11,526, doi:10.1029/JB089iB13p11517.

Milano, G., G. Ventura, and R. Di Giovambattista (2002), Seismic evi-dence of longitudinal extension in the southern Apennines chain (Italy):The 1997–1998 Sannio‐Matese seismic sequence, Geophys. Res. Lett.,29(20), 2004, doi:10.1029/2002GL015188.

Miller, S. A., C. Collettini, L. Chiaraluce, M. Cocco, M. Barchi, and B. J.P. Kaus (2004), Aftershocks driven by a high‐pressure CO2 source atdepth, Nature, 427, 724–727, doi:10.1038/nature02251.

Montone, P., M. T. Mariucci, S. Pondrelli, and A. Amato (2004), Animproved stress map for Italy and surrounding regions (central Mediter-ranean), J. Geophys. Res., 109, B10410, doi:10.1029/2003JB002703.

Moos, D., and M. D. Zoback (1983), In situ studies of velocity in fracturedcrystalline rocks, J. Geophys. Res., 88, 2345–2358, doi:10.1029/JB088iB03p02345.

Moretti, M., P. De Gori, and C. Chiarabba (2009), Earthquake relocationand three‐dimensional Vp and Vp/Vs models along the low angle AltoTiberina fault (central Italy): Evidence for fluid overpressure, Geophys.J. Int., 176, 833–846, doi:10.1111/j.1365-246X.2008.03984.x.

Morris, A., D. A. Ferrill, and D. B. Henderson (1996), Slip‐tendency anal-ysis and fault reactivation, Geology, 24, 275–278, doi:10.1130/0091-7613(1996)024<0275:STAAFR>2.3.CO;2.

Nadeau, R., M. Antolik, P. Johnson, W. Foxall, and T. V. McEvilly (1994),Seismological studies at Parkfield III: Microearthquake clusters in thestudy of fault‐zone dynamics, Bull. Seismol. Soc. Am., 83, 247–263.

Noir, J., E. Jacques, S. Bekri, P. M. Adler, P. Tapponier, and G. C. P. King(1997), Fluid flow triggered migration of events in the 1989 Dobi earth-quake sequence of central Afar, Geophys. Res. Lett., 24, 2335–2338,doi:10.1029/97GL02182.

Nostro, C., M. Cocco, and M. E. Belardinelli (1997), Static stress changesin extensional regimes: An application to southern Apennines, Italy, Bull.Seismol. Soc. Am., 87, 234–248.

Patacca, E., P. Scandone, E. Di Luzio, G. Cavinato, and M. Parotto (2008),Structural architecture of the central Apennines: Interpretation of theCROP 11 seismic profile from the Adriatic coast to the orographicdivide, Tectonics, 27, TC3006, doi:10.1029/2005TC001917.

Peacock, D. C. P. (2002), Propagation, interaction and linkage in normalfault systems, Earth Sci. Rev., 58(1–2), 121–142, doi:10.1016/S0012-8252(01)00085-X.

Pino, N. A., and F. Di Luccio (2009), Source complexity of the 6 April2009 L’Aquila (central Italy) earthquake and its strongest aftershockrevealed by elementary seismological analysis, Geophys. Res. Lett.,36, L23305, doi:10.1029/2009GL041331.

Pizzi, A., and F. Galadini (2009), Pre‐existing cross‐structures and activefault segmentation in the northern‐central Apennines (Italy), Tectonophy-sics, 476, 304–319, doi:10.1016/j.tecto.2009.03.018.

Roeloffs, E. (1996), Poroelastic techniques in the study of earthquakerelated hydrologic phenomena, Adv. Geophys., 37, 136–195.

Saccorotti, G., G. Ventura, and G. Vilardo (2002), Seismic swarms relatedto diffusive processes: The case of Somma‐Vesuvius volcano, Italy, Geo-physics, 67(1), 199–203, doi:10.1190/1.1451551.

Satolli, S., and F. Calamita (2008), Differences and similarities between thecentral and the southern Apennines (Italy): Examining the Gran Sassoversus the Matese‐Frosolone salients using paleomagnetic, geological,

and structural data, J. Geophys. Res., 113, B10101, doi:10.1029/2008JB005699.

Shapiro, S. A., E. Huenges, and G. Borm (1997), Estimating the crustpermeability from fluid‐injection‐induced seismic emission at the KTBsite, Geophys. J. Int., 131, F15–F18, doi:10.1111/j.1365-246X.1997.tb01215.x.

Shapiro, S. A., R. Patzig, E. Rothert, and J. Rindschwentner (2003), Trig-gering of seismicity by pore‐pressure perturbations: Permeability‐relatedsignatures of the phenomenon, Pure Appl. Geophys., 160, 1051–1066,doi:10.1007/PL00012560.

Sibson, R. H. (2000), Fluid involvement in normal faulting, J. Geodyn., 29,469–499, doi:10.1016/S0264-3707(99)00042-3.

Stern, R. J. (2002), Subduction zones, Rev. Geophys., 40(4), 1012,doi:10.1029/2001RG000108.

Talwani, P., L. Chen, and K. Gahalaut (2007), Seismogenic permeability,ks, J. Geophys. Res., 112, B07309, doi:10.1029/2006JB004665.

Tertulliani, A., A. Rossi, L. Cucci, and M. Vecchi (2009), L’Aquila (centralItaly) earthquakes: The predecessors of the April 6, 2009 event, Seismol.Res. Lett., 80(6), 1008–1013, doi:10.1785/gssrl.80.6.1008.

Ventura, G., F. R. Cinti, F. Di Luccio, and N. A. Pino (2007), Mantlewedge dynamics versus crustal seismicity in the Apennines (Italy),Geochem. Geophys. Geosyst., 8, Q02013, doi:10.1029/2006GC001421.

Wadati, K. (1933), On the travel time of earthquake waves. Part II,Geophys.Mag., 7, 101–111.

Waldhauser, F., and W. L. Ellsworth (2000), A double‐difference earth-quake location algorithm: Method and application to the northernHayward fault, California, Bull. Seismol. Soc. Am., 90, 1353–1368,doi:10.1785/0120000006.

Wessel, P., and W. H. F. Smith (1991), Free software helps map anddisplay data, Eos Trans. AGU, 72, 441, doi:10.1029/90EO00319.

Yin, Z. M., and G. Ranalli (1992), Critical stress difference, fault orienta-tion and slip direction in anisotropic rocks under non‐Andersonian stresssystems, J. Struct. Geol., 14, 237–244, doi:10.1016/0191-8141(92)90060-A.

Zhang, T., M. Zhang, B. Bai, X. Wang, and L. Li (2008), Origin and accu-mulation of carbon dioxide in the Huanghua depression, Bohai BayBasin, China, AAPG Bull., 92, 341–358, doi:10.1306/10230706141.

Zhang, X., and D. J. Sanderson (1996a), Numerical modelling of the effectsof fault slip on fluid flow around extensional faults, J. Struct. Geol., 18,109–119, doi:10.1016/0191-8141(95)00086-S.

Zhang, X., and D. J. Sanderson (1996b), Effects of stress on the two‐dimensional permeability tensor of natural fracture networks, Geophys.J. Int., 125, 912–924, doi:10.1111/j.1365-246X.1996.tb06034.x.

Zhao, D., and H. Negishi (1998), The 1995 Kobe earthquake: Seismicimage of the source zone and its implications for the rupture nucleation,J. Geophys. Res., 103, 9967–9986, doi:10.1029/97JB03670.

Zhao, D., H. Kanamori, H. Negishi, and D. Wiens (1996), Tomography ofthe source area of the 1995 Kobe earthquake: Evidence for fluids at thehypocenter?, Science, 274, 1891–1894, doi:10.1126/science.274.5294.1891.

F. R. Cinti, F. Di Luccio, and G. Ventura, Istituto Nazionale di Geofisicae Vulcanologia, Sismologia e Tettonofisica, Via di Vigna Murata 605,I‐00143 Roma, Italy. (francesca.diluccio@ingv.it)R. Di Giovambattista and A. Piscini, Istituto Nazionale di Geofisica e

Vulcanologia, Centro Nazionale Terremoti, Via di Vigna Murata 605,I‐00143 Roma, Italy.

DI LUCCIO ET AL.: THE 2009 L’AQUILA EARTHQUAKE, CENTRAL ITALY B06315B06315

15 of 15