Three-dimensional segmentation and different rupture behaviour during the 2012 Emilia seismic sequence (Northern Italy)

�������� ����� ��

Three-dimensional segmentation and different rupture behaviour during the2012 Emilia seismic sequence (Northern Italy)

Lorenzo Bonini, Giovanni Toscani, Silvio Seno

PII: S0040-1951(14)00243-1DOI: doi: 10.1016/j.tecto.2014.05.006Reference: TECTO 126307

To appear in: Tectonophysics

Received date: 8 October 2013Revised date: 28 April 2014Accepted date: 5 May 2014

Please cite this article as: Bonini, Lorenzo, Toscani, Giovanni, Seno, Silvio, Three-dimensional segmentation and different rupture behaviour during the 2012 Emilia seismicsequence (Northern Italy), Tectonophysics (2014), doi: 10.1016/j.tecto.2014.05.006

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

1

Three-dimensional segmentation and different rupture behaviour

during the 2012 Emilia seismic sequence (Northern Italy)

Lorenzo Bonini1*

, Giovanni Toscani1, Silvio Seno

1

1 Dipartimento di Scienze della Terra e dell’Ambiente, Università di Pavia, via Ferrata 1, 27100

Pavia, Italy

*Corresponding author: Lorenzo Bonini, e-mail address: [email protected]

Abstract

Describing the slip behaviour of an active fault system is central to understanding seismic potential

of seismogenic areas. Different elements control the nature and the extent of the coseismic and post-

seismic ruptures, including the geometry of faults, the nature of faulted rocks, and the stress

changes caused by the mainshocks. In May-June 2012 a severe seismic sequence struck a portion of

the Po Plain (Northern Italy), where a thick blanket of Plio-Quaternary sediments hides a number of

seismogenic sources corresponding to the external thrust systems of the Northern Apennines. We

used deep seismic reflection data to reconstruct the geometry of the faults responsible for the

sequence. These faults exhibit significant non-planarity due to tectono-stratigraphic heterogeneities

inherited from a complex pre-thrusting extensional tectonic phase. A comparison of the fault

parameters derived from our geological analysis and the evidence supplied by seismological

(aftershock distributions) and geodetic data (InSAR) allowed to identify the causative fault

segments of the two mainshocks. We then modeled the Coulomb stress changes produced by two

mainshocks to analyze on- and off-fault seismicity. Discrepancies between the magnitude of the

earthquakes and the extent of active faults suggest that the mainshocks did not rupture the entire

thrust planes. We contend that seismogenic ruptures were confined in the Mesozoic carbonates and

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

2

were stopped by lithological changes and/or mechanical complexities of the fault planes, both along

dip and along strike. Our findings highlight that along the active structures of the Po Plain slip tends

to be seismogenic where faults are located in Mesozoic carbonate rocks.

Keywords:

Thrust faults

Slip behaviour

Fault segmentation

Po Plain

The 2012 Emilia earthquakes

1. Introduction

Since the advent of modern seismology, the distribution of coseismic slip on the presumed fault

plane has been used to image earthquake ruptures (e.g.: Haskell, 1964). During the past three

decades, new geodetic tools (GPS networks, Synthetic Aperture Radar) have become available: the

data have been promptly introduced into joint inversion models to constrain the final fault slip,

providing a fundamental tool to investigate the behavior of active fault zones. Notwithstanding the

significant progress achieved in recent years, our capability to detect complex geometries using

these methods, however, is often below the resolution of the data they are based on.

It is known that the understanding the relationships between the architecture of the fault-

zone and the earthquake potential is central to investigating the slip behavior in an active fault

systems (e.g., Marone and Richardson, 2010). Strong constraints on the location and geometry of

active fault segments are therefore a fundamental prerequisite for improving our understanding of

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

3

the physics of earthquakes. The cross-correlation between seismological data (e.g.: aftershock

locations) and seismic reflection data is one of the most attractive and useful techniques to image

faults activated at depth (e.g.: Carena and Suppe, 2002; Fuis et al., 2003). Such practice, however, is

not commonly used as seismic reflection data are seldom made freely available by the oil industry.

Over the past 70 years extensive hydrocarbon exploration has shed light on a number of

buried thrust-faults belonging to the external parts of the Northern Apennines (Italy) and currently

buried beneath the adjacent Po Plain, (Pieri and Groppi, 1981; Cassano et al., 1986; Fantoni and

Franciosi, 2010). Some of these structures are known to be seismically active, as shown by the

severe seismic sequence that struck the Emilia region with two mainshocks on 20 and 29 May 2012

(MW 6.1 and 6.0, respectively: Regional Centroid Moment Tensor, RCMT;

http://www.bo.ingv.it/RCMT; Fig. 1). Although some of these events were generated by previously

identified active faults (e.g.: Burrato et al., 2003; Toscani et al., 2009; DISS Working Group, 2010),

the availability of seismic reflection data in the epicentral area now affords the rare opportunity to

improve the characterization of those buried faults by merging and comparing geodetic and

seismological data with seismic images of the causative faults at depth.

In this work, we correlate seismological data (i.e., mainshock and aftershock locations as

well as focal mechanisms) with sub-surface geological structures. Based on seismological and

geological evidence we identify the fault segments that ruptured during the mainshocks and

compute changes in static Coulomb stress surrounding the main coseismic patches to analyze the

aftershock distribution. We show that the spatial distribution of ruptures during the 2012 Emilia

sequence was controlled by three-dimensional segmentation driven by local geological conditions,

i.e. by the non-planar geometry of the active faults and the by variability of slip behavior (stable or

unstable slip) due to the distribution of rock types.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

4

2. The Po Plain and the 2012 Emilia sequence

The Apennines fold-and-thrust belt evolved in the framework of the Africa-Europe convergence

since the Cretaceous and developed mainly during Neogene and Quaternary (Malinverno and Ryan,

1986, Royden at al., 1987; Dewey et al., 1989). The northern Apennines, in particular, are a NNE-

verging thrust belt, partially exposed in its inner parts along the mountain range and buried in its

outer parts by the alluvial sediments of the Po Plain, where the belt is shaped in arched thrust

systems that exhibit NNE-ward convexity. Evidence of ongoing compressional tectonics can be

directly observed in the Apennines foothills, whereas the core of the chain is currently undergoing

extension (Bertotti et al., 1997; Picotti and Pazzaglia, 2008).

Three main arcs compose the buried part of the Northern Apennines (Fig. 1); they formed at

different times and exhibit variable shortening rates. From W to E they are: (i) the Monferrato arc,

whose inception is dated to the Messinian (Bertotti and Mosca, 2009); (ii) the Emilia arc, mainly

developed from Messinian to Late Pleistocene (Toscani et al., 2006); and (iii) the Ferrara arc, active

since Late Pliocene (Scrocca et al., 2007; Toscani et al., 2009; Boccaletti et al., 2011; Ahmad et al.,

2013). All three arcs are strongly asymmetrical, the eastern lateral ramps being more developed

than the western ones (Pieri and Groppi, 1981; Castellarin and Vai, 1986; Patacca et al., 1992). As a

whole, shortening along the thrust fronts increases from the Monferrato to the Ferrara arc, i.e., from

WNW to ESE, due to a counterclockwise rotation of the entire fold-and-thrust belt that occurred

during its emplacement (Vanossi et al., 1994; Muttoni et al., 2001; Maino et al., 2013).

The foreland basin of the NE-verging Northern Apennines corresponds to the Po Plain, which

separates the Apennines foothills, to the south, from the Southern Alps mountain range, to the

north. Since Late Cretaceous, the Po Plain area has acted as the foreland of both these opposite-

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

5

verging chains, which developed as a result of the continental collision between the European and

the African Plates (Robertson and Grasso, 1995; Boccaletti et al., 2004).

During the Triassic and Jurassic, the region was affected by an extensional tectonic phase

(known as Mesozoic rifting phase; Bertotti et al., 1993, Fantoni and Franciosi, 2010) that caused

large stratigraphic contrasts in the sedimentary successions. The subsequent Late Cenozoic

evolution of the thrust systems was strongly controlled by thickness variations in the stratigraphic

succession, which often correspond to tectonic-controlled abrupt rheological changes (Ravaglia et

al., 2004; 2006).

Summarizing, the sedimentary infill of the Po Plain and the underlying tectonostratigraphic

units provide a full record of the tectonic phases that led to the current structural setting, from the

Triassic-Jurassic passive margin conditions to the Cretaceous-to-present compressional regime,

which includes the formation of a Neogene-Quaternary foredeep basin (e.g.: Dondi and D’Andrea,

1986; Argnani and Ricci Lucchi, 2001; Ghielmi et al., 2013).

The tectonic setting of the buried Ferrara arc is characterized by thrust faults with relatively

shallow thrust anticlines and deeper and wider synclines (Pieri and Groppi, 1981; Cassano et al.,

1986; Bigi et al., 1990). Shallow well logs and high resolution seismic lines allow the architecture

of the most recent deposits to be described in detail (R.E.R. and ENI-AGIP, 1998). The integration

of these data contributes to define a system of NE-verging blind thrusts and related folds; these

control the syntectonic deposition of sedimentary wedges, whose thickness is highly variable,

locally reaching up to 7-8 km. A complex pattern of Mesozoic normal faults affects the oldest

stratigraphic levels of the Ferrara arc (Fig. 1). These faults segmented the entire Po plain foreland,

which is now deflected below the frontal orogenic wedge of the Northern Apennines, and have been

partly reactivated by the subsequent compressional tectonic regime.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

6

As mentioned earlier, the Northern Apennines fold-and-thrust belt and the related foredeeps

migrated northeastwards during the Late Oligocene-Quaternary time interval (Malinverno and

Ryan, 1986; Royden et al., 1987; Doglioni, 1991; Scrocca et al., 2007), and are still active. The

central part of the Po Plain, a low relief area, does not exhibit any remarkable surface evidence of

the ongoing growth of the buried anticlines. The differential vertical motions caused by their recent

and current activity, however, controls the drainage pattern, leading to river diversions and changes

in channel patterns. These morphotectonic elements have been recognized and mapped by Burrato

et al. (2003), who suggested the presence of active thrusts in the subsurface and interpreted them as

potential seismogenic sources (DISS Working Group, 2010, with references).

The 2012 Emilia earthquakes struck an area historically affected by low-to-moderate size

earthquakes (M<6.0: from CPTI11 catalogue, Rovida et al., 2011). The best known historical

earthquake that occurred in the study area is the long seismic sequence which started on 17

November 1570 near Ferrara with four strong shocks (reported magnitude of the largest is Mw 5.4)

that caused severe structural damage and partial collapses in Ferrara and its surroundings

(Guidoboni et al., 2007). The sequence went on discontinuously up to March 17, 1574, when the

last significant shock is reported to have caused damage in the village of Finale Emilia, about 25 km

west of Ferrara.

2.1. The 2012 Emilia earthquake sequence

The 2012 Emilia sequence started on May 20 (02:03:53 UTC; MW 6.1, ML 5.9). It was preceded by

five foreshocks (Scognamiglio et al., 2012) the most energetic of which was a ML 4.1 event that

occurred two hours before the 20 May mainshock. A ML 5.1 event occurred five minutes after the

mainshock (02:07:31 UTC), starting a long sequence punctuated by rather large aftershocks. From

May 20 to May 29, several earthquakes spread over a ~30 km long and ~15 km wide area

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

7

(http://iside.rm.ingv.it; Fig. 1). On May 29, a second large shock (07:00:03 UTC; MW 6.0, ML 5.8),

struck an adjacent area to the west of the first mainshock, followed by two strong aftershocks

having ML 5.3 (10:55:57 UTC) and ML 5.2 (11:00:25 UTC), respectively. The last ML ˃5 event of

the sequence occurred on 3 June (ML 5.1). Overall, about 2,500 events were recorded by permanent

and temporary INGV seismometers during the first month, of which 30 had ML ˃4

(http://iside.rm.ingv.it). The focal mechanisms of the two mainshocks as well as those of the largest

aftershocks showed reverse slip kinematics on ca. E-W-striking nodal planes (Pondrelli et al., 2012;

Saraò and Peruzza, 2012; Scognamiglio et al., 2012), in agreement with GPS data (e.g. Devoti et al.,

2011), borehole breakouts (Montone et al., 1999), centroid moment tensor solutions (Pondrelli et

al., 2006) and with the present-day structural setting of the area (Fig. 1). The focal depths of the two

mainshocks were estimated 6.3 km for the May 20 event and 10.2 km for the May 29 event,

respectively (http://iside.rm.ingv.it); all the other shocks are distributed in the depth interval 4 to 15

km.

DInSAR measurements based on extensive Radarsat-1, Radarsat-2 and COSMO-SkyMed

datasets constrained by scattered GPS observations revealed sizable coseismic crustal strains. The

deformation pattern consists of two en echelon, gently-asymmetric, ~25 km-long and ~10 km wides

upside down bowls peaking at 15-21 cm and trending WNW-ESE (e.g.: Pezzo et al., 2013; Tizzani

et al., 2013).

3. Buried faults: evidence from seismic reflection data

The VIDEPI database (http://unmig.sviluppoeconomico.gov.it/videpi/en/) makes publicly available

a number of seismic lines crisscrossing the entire Po Plain, plus many borehole data. In order to

image the causative faults of the 2012 sequence we selected a deep reflection profile (BO-365-88)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

8

crossing the Ferrara arc close to the epicenters of the 20 and 29 May earthquakes (Fig. 1). The 60

km-long profile is rectilinear and strikes N22°E (Fig. 2a). We analyzed both its stack and migrated

versions. The profiles are cut at 7,000 ms and 6,000 ms TWT (Two Way Time), respectively, and

their joint study allowed us to investigate the whole stratigraphic succession from the Variscan

basement up to Quaternary sediments. We used wireline logs, descriptions of cores and cuttings

from 5 wells (Pilastri 1, Rivara 1, Camurana 2, Spada 1 and Bignardi 1; available at:

http://unmig.sviluppoeconomico.gov.it/videpi/en/) to calibrate and depth-convert the seismic

profile. Finally, to validate our interpretation we restored the obtained geological section following

classical, well established rules (e.g.: Dahlstrom, 1969; Elliott, 1982).

Our interpretation displays four main groups of NNE-verging, thrust-ramp anticlines (Fig.

2b). The innermost group (Castelfranco Emilia and Nonantola), located at the SSW end of the

section, is composed by three fore-thrusts laid one on top of the other at different structural levels.

The shallower thrust is rooted into the Upper Eocene-Upper Miocene deposits and affects the Upper

Eocene-Upper Miocene successions, which end with regressive Messinian evaporites and Lower

Pliocene sediments. Two deeper thrusts displace the Upper Triassic and Lower Jurassic carbonate

succession up to the Upper Jurassic – Eocene deposits.

In the central part of the cross-section, a second group of thrust-ramp anticlines is formed by a

well visible and developed thrust system (Mirandola; Fig. 2b). To the SSW, this structure is rooted

into a basal décollement displaying flat attitude and located in Lower–Middle Triassic rocks

consisting of continental siliciclastic deposits and evaporites. A sudden change in its dip angle

marks the transition to the thrust-ramp anticline, which displays fore- and back-thrusts cutting the

Meso-Cenozoic rocks and deforming the Quaternary deposits.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

9

At a depth of 5.5-6.0 s TWT in the seismic reflection profile (Fig. 2a), which corresponds to a

depth of about 12 km in the depth-converted cross-section (Fig. 2b), the Mirandola thrust ramp is

notably superimposed onto deeper structural elements. Here the reflectors exhibit fan geometries

revealing the presence of SSW-dipping, high angle normal faults clearly displacing Carnian and

Lower Liassic deposits (Fig. 2a, inset). Moving upwards, this fan is topped by a set of

discontinuous, high amplitude reflectors. We associated this seismic facies with the top of the

Variscan basement (Fantoni and Franciosi, 2010), that can be recognized all along the seismic

profile; apart from minor reverse displacements seen near the thrusts, this marker dips gently to the

SW in keeping with the general dip of the Apennines monocline underneath the Ferrara arc

(Mariotti and Doglioni, 2000).

In the shallowest part of the seismic line, between 0.0 and 3.0 s TWT, lying above the

previously described normal faults the Mirandola ramp anticline brings the Meso-Cenozoic

sedimentary successions at higher structural levels, such that two oil wells drilled less than 5 km

from the trace of the seismic profile encountered Jurassic deposits at 3,400 m depth. As shown also

by other investigators (e.g. Scrocca et al., 2007), the bottom of the overlying Plio-Quaternary

deposits is locally displaced by the upper termination of the splays connected to the thrust ramp,

whose recent-to-current activity is further testified by growth strata and onlap geometries clearly

visible in the shallowest part of the profile (between 0.0 and 1.0 s TWT).

A third group of thrust-ramp anticlines (San Martino and Pilastri; Fig. 2b) is seen as one

proceeds towards the NNE. It is formed by two parallel, deeply rooted thrusts that have been

interpreted in the uppermost 5 s TWT of the seismic profile and extrapolated downward. At shallow

depth, both thrusts displace the Pliocene-Quaternary deposits that in some cases form a very thin

cover on the ramp anticlines and show fore- and back-limb onlap geometries.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

10

The fourth and northernmost group of structures surveyed is formed by a thrust-backthrust

pop-up, whose top is slightly deeper than the previous ones. The activity of this tectonic structure

can be dated to the Pliocene-Quaternary, as witnessed by gently folded reflectors ascribed to this

time interval.

The cross-section restoration of Figure 2c aims at retrieving the pre-compressional stage and

shedding light on the kinematics of faulting. The proposed pre-thrusting reconstruction shows a

complex system of extensional structures. The structural control exerted by inherited normal faults

on the evolution of the subsequent thrusting phase is a widely recognized feature of the Apennines,

where (i) cases of positive inversion, (ii) normal faults truncated by younger thrusts, and (iii)

normal faults providing stress concentration and promoting thrust localization have all been

described (e.g.: Scisciani, 2009; Di Domenica et al., 2014 and references therein). Figure 2c shows

the trajectories of the thrust faults, allowing us to describe the structural evolution of the identified

thrusts.

Starting from the Mirandola structure, our interpretation of the seismic section shows that this

thrust started as a décollement within Lower-Middle Triassic units and stopped its horizontal

propagation against a pre-existing extensional fault. This inherited fault probably provided stress

concentration and promoted the subsequent propagation of a thrust-ramp in the Mesozoic carbonate

rocks. The newly formed reverse fault cut both the hangingwall and the footwall of the inherited

extensional fault (footwall shortcut thrust). As a consequence, the upper portion of the previous

extensional system was incorporated in the Mirandola thrust sheet. Hence, some of the secondary

faults located at the core of the Mirandola anticline probably derive from the reactivation of

inherited extensional segments. Such hypothesis could explain the differences in structural style

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

11

among the Mirandola structure and the other external thrusts, namely the San Martino and Pilastri

structures. Indeed, these two thrust-faults appear as result of a standard flat-ramp-flat evolution.

With regard to the most external thrust – the Ficarolo structure – the high angle fault is more

consistent with the reactivation of an inherited extensional fault than with the evolution of a reverse

fault. This interpretation is supported by the different dip of other reverse faults that developed in

the same rock types. For instance, the San Martino and Pilastri structures show a fault angle ranging

from 30° to 35°, i.e., much gentler than the 55°-60° dip of the Ficarolo fault.

4. The 2012 Emilia sequence: seismicity and geology

The 2012 seismic sequence is well illustrated by a catalogue of 1,104 relocated events supplied by

Marzorati et al. (2012). When coupled with our geological cross section, these data provide a

valuable contribution to the identification of the individual faults activated during the sequence

(Fig. 3). We plotted separately the aftershocks following the first (20 May, Mw 6.1) and the second

mainshock (29 May, Mw 6.0). Thus, we obtained two clusters of seismicity. The first cluster groups

aftershocks from May 20 to May 28 and shows events that align rather regularly along the San

Martino thrust. The second cluster encompasses earthquakes starting from the 29 May onward.

Their distribution is more scattered, though the aftershocks are basically centered around the

Mirandola structure. Hence the location of the mainshocks and the aftershocks in itself indicates the

San Martino and Mirandola thrusts as the causative faults of the two earthquake sequences.

Analytic and numerical inversion models based on InSAR observations produced by

different investigators (Bignami et al., 2012; Pezzo et al., 2013; Tizzani et al., 2013) confirmed

previous speculations based on aftershock data. All published models regarding the 20 May

mainshock agree that the coseismic rupture occurred between 4 to 8 km depth along the San

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

12

Martino thrust (Fig. 4c). As for the source of the 29 May shock, the inversion models indicate

coseismic slip along the Mirandola structure between 3 and 10 km depth (Fig. 4c).

The Mirandola and San Martino thrusts are known to be over 100 km-long (e.g. Bigi et al.,

1990) and the cross-section indicates they are 15-20 km-wide along dip (Figs. 1 and 3). Empirical

relationships between earthquake magnitude and rupture area (e.g. Wells and Coppersmith, 1994)

suggest that during the Emilia sequence the seismogenic ruptures did not involve the entire thrust

fault planes. When considering the aftershock distribution with respect to the lithologies that can be

found at different focal depths, one can observe that most earthquakes nucleated within the

Mesozoic carbonate rocks, whereas only few events are located within the Cenozoic turbiditic and

clastic sediments (Fig. 3b). Such behavior becomes even more evident when plotting the

lithological setting proposed in our geological section onto the coseismic slip model obtained by

Ganas et al. (2012) for the San Martino thrust plane, the causative source of the 20 May event (Fig.

5a). Coseismic slip is confined within the Upper Triassic dolostones and Lower Jurassic limestones;

in particular, the upper cutoff corresponds to the boundary between Lower Jurassic limestones and

Cenozoic sediments, whereas the lower cutoff is located close to the transition between Lower-

Middle Triassic deposits and Upper Triassic dolostones.

Combining the slip inversion results derived from geodetic data by Pezzo et al. (2013) along

San Martino and Mirandola thrusts with our data we confirm the previous observation, stressing

that coseismic slip areas are located in the Mesozoic carbonate rocks (Figs 5b and 5c). We conclude

that the major seismogenic ruptures of our study area are located in Mesozoic carbonates and that

Cenozoic sediments have a velocity strengthening behaviour.

In addition to the observed correlation between the earthquake distribution and the rock

types that occur at depth along the seismic-aseismic transition, we also remark that the seismogenic

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

13

segment boundaries also correspond to changes in the fault geometry, where the upper and lower

seismicity cut-offs coincide with the change of the fault plane dip - from ramp to flat and vice versa,

both for the San Martino and Mirandola thrusts.

4.1. Stress transfer and aftershocks pattern distribution

Any interpretation of fault geometry exclusively based on aftershock clustering is inherently

ambiguous. Some aftershocks are concentrated at the edge of slipped patches, or caused by steps

and jogs of the fault plane, or fall at fault intersections (on-fault seismicity); others may be related

to the reactivation of faults located near the main faults (off-fault seismicity). Aftershocks located at

the edge of the main slipped patches and off-fault seismicity can be separated evaluating the stress

perturbation produced by main ruptures.

Spatial correlation of aftershock hypocenters and coseismic stress changes has been widely

used to analyze seismic sequences (e.g. Stein et al., 1994; Harris et al., 1995; Wang and Chen,

2001; Lin and Stein, 2004). Having derived from our geological section a fault geometry that is

both reliable and independent from aftershock clustering, we can investigate the seismicity

distribution separating on- and off-fault seismicity. Stress change calculations for the 2012 Emilia

sequence have been performed also by other workers (e.g., Ganas et al., 2012; Sarao and Peruzza,

2012; Pezzo et al., 2013), but they mainly focused on the relation between the two principal shocks,

confirming that the 20 May mainshock caused a stress increase on the causative fault of the 29 May

mainshock.

In its simplest form, the static stress change ΔCFF (i.e. the stress variation described by the

Coulomb Failure Function) produced by an earthquake on a nearby fault is given by:

ΔCFF=Δτ+μ’Δσn

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

14

where Δτ is the shear stress change on a given fault plane, μ’ is the effective coefficient of friction,

and Δσn is the normal stress change (positive if the fault is unclamped).

We modeled the ΔCFF on optimally oriented faults under the assumption that all faults are

embedded in an elastic, homogeneous and isotropic half-space with μ’ = 0.8 as the effective

coefficient of friction. All geometric parameters and the slip distribution were taken from the work

of Pezzo et al. (2013). Available focal mechanisms provide rakes of 90° for the 20 May and 85° for

the 29 May shocks (Malagnini et al., 2012). All calculations were performed using the Coulomb 3.3

software (Lin and Stein, 2004; Toda et al., 2005).

The ΔCFF caused by the 20 May event (Fig. 6a) produced four lobes of Coulomb stress rise

(red areas) and four lobes of stress drop (blue areas). The lobes at the upper and lower end of the

considered segment highlight crustal volumes that were made more prone to suffer additional on-

fault seismicity as they fall at the edge of main slipped patch. No aftershocks occur in the upper

portion of the activated thrust, in accordance with the suggested aseismic behavior of the fault

segments located in the Quaternary sediments. Most of the aftershocks are located along the low-

angle, deeper segment of the San Martino thrust, due to slip on this portion of the thrust and to

partial reactivation of other inherited structures. Stress increase also affects the lower portion of the

Pilastri structure. No aftershocks are associated with this structure, however, probably because the

ΔCFF produced by the 20 May event was relatively small. A stress increase is seen also in the

uppermost part of the Mirandola structure, but no aftershocks are seen in this rock volume as it is

assumed to be dominantly aseismic.

Regarding the 29 May event (Fig. 6b), the coseismic slip induced a stress increase at the

edge of the ruptured area on the Mirandola structure. A dense cluster of aftershocks is seen centered

at the core of the anticline, that is formed by Mesozoic carbonates. No aftershocks are located along

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

15

the lower and flat portion of the thrust, suggesting aseismic behavior. The scattered aftershocks

located below the Mirandola anticline may correspond to the activation of both minor faults in the

San Martino thrust sheet and of an inherited structure located in its footwall. In this view, the

vertical alignment of the aftershocks related to the 29 May event does not correspond to a single

structural feature, because it is related to (1) earthquakes located on secondary faults of the

Mirandola anticline and directly connected with the master fault, and (2) to the off-fault seismicity

produced by the activation of minor faults.

4.2. Lateral segmentation of seismogenic sources

Regarding the lateral extent of the ruptured areas, the number of the deep seismic reflection profiles

publicly available is not sufficient to reconstruct an accurate three-dimensional image of the fault

systems in the whole Ferrara arc. Nevertheless, a structural map of the top of Mesozoic carbonates

derived from shallow seismic profiles is available for the eastern part of the Mirandola structure

(Nardon et al., 1990; Figure 7). This image suggests that the most energetic aftershocks (M˃5)

following the second mainshock fall along the lateral extension of the causative thrust of the 29

May earthquake. This implies that seismogenic ruptures may be larger than those seen in May 2012.

Whether or not the 2012 ruptures represent the characteristic earthquakes for this area depends on

the nature of the geological features that could act as barriers. It is well known that aseismic (stable)

zones and mechanical discontinuities (e.g. pre-existing faults or weak layers) may act as an

obstacle, hence stopping or somehow confining the seismogenic ruptures (e.g. Sibson, 1985;

Roering et al., 1997; Scholz, 1998). On the one hand, we know that aseismic areas may be involved

in coseismic slip depending on the amount of coseismic slip (e.g. Noda and Lapusta, 2013). The

moderate magnitude of the 2012 Emilia earthquakes (M~6) makes this hypothesis unlikely because

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

16

the maximum seismogenic slip (60 cm; Pezzo et al., 2013) is too small to generate coseismic sliding

in stable regions (see Noda and Lapusta, 2013 for a discussion). On the other hand, the arrest of

earthquake ruptures at mechanical discontinuities depends on the thickness of this weaker or

stronger layer, on its orientation compared with the geometry of the seismogenic fault, and on the

role played by fluids.

Figure 7 shows that the Mirandola thrust plane is laterally segmented by inherited

extensional faults incorporated by the thrust sheet during the compressional phase (see section 3).

Hence, the lateral extent of the seismogenic rupture can be expected to be limited by these inherited

structures. Though we do not have a high-resolution three-dimensional image of the San Martino

thrust, the lateral migration of seismicity recorded during the 20 May sequence suggests that lateral

segmentation of this source is likely. If this is the case, the seismogenic barriers in the Emilia area

could have a persistent character, making the 2012 earthquakes characteristic for the area, in

agreement with historical catalogue (Rovida et al., 2011) and previous estimates based on

geological observations and hypotheses (Valensise and Pantosti, 2001; Burrato et al., 2003; DISS

Working Group, 2010).

5. Conclusions

The interpretation of the seismic profile located across the epicentral area, the analysis of the

aftershocks distribution at depth, and the investigation of the static stress changes produced by the

two mainshocks, all of this allowed us to propose a new seismotectonic model that is in full

agreement with all the data available on the 2012 Emilia sequence. Figure 8 summarizes the

seismotectonic model, showing the seismogenic and aseismic parts of the two activated thrusts.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

17

We infer that the mainshock ruptures nucleated in Mesozoic carbonates and were stopped by

lithological and structural variations along the activated faults, such as (i) changes in fault plane

geometry from ramp to flat and viceversa, and (ii) changes in the lithostratigraphic succession. The

lack of shallow aftershocks, coupled with the available fault displacement models, suggests that no

seismic slip occurred along the secondary faults cutting Quaternary deposits. Nonetheless, aseismic

slip is likely to occur along these segments during the mid- to long-term postseismic phase (i.e.

mid-to long-term afterslip).

In summary, we propose that the segmentation of the 20 and 29 May seismogenic sources

can be explained as related to: 1) the change in frictional behaviour during slip, due to the variations

of the mechanical stratigraphy (from velocity weakening to velocity strengthening; e.g. Scholz,

1998); 2) the variations of the fault geometry and the occurrence of secondary faults, which stopped

the dynamic rupture, according to another well known segmentation mechanism (e.g. Poliakov et

al., 2002; Bonini et al., 2014). Concerning the deep and flat portion of the Mirandola thrust, we

believe that the lack of aftershocks aligned along this structure is consistent with the occurrence of

aseismic creeping, whereas the aftershocks located at the footwall of the activated thrusts may

imply a partial inversion of pre-existing extensional faults. Such occurrence has also been

documented in other seismotectonic settings, for instance the reactivation of normal faults at the

footwall of megathrusts in subduction zones (e.g. Nakajima et al., 2011).

Our findings have important implications for a more detailed characterization of the seismic

hazard in the Po plain area, one the most densely populated areas of Europe. Hence, we can use

what we learned from the 2012 earthquakes to make some speculations on the seismic behavior of

other active faults in this region. According to the Italian earthquake catalogue (Rovida et al., 2011)

and to the Database of Individual Seismogenic Source of Italy (DISS: http://diss.rm.ingv.it/diss/),

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

18

several other faults are capable of generating earthquakes of magnitude up to 6.7 in the Po Plain

(e.g. the 1117 Verona earthquake, Mw 6.7; Guidoboni et al., 2007). The causative faults of these

events are located along the external thrusts of the Southern Alps and Northern Apennines, in

seismotectonic scenarios that are similar to the that of the Ferrara arc. The earthquake potential of

the Po Plain faults is very uncertain, however, and the limited earthquake record suggests a rather

low seismic coupling, i.e. the ratio between the observed seismic moment release and the potential

calculated based on the total size of the identified faults. As suggested by our study, the actual

seismogenic layers may be confined in the Mesozoic carbonate rocks and large portions of the

thrusts may exhibit aseismic behavior. Hence our findings may help assessing the seismogenic

potential of other active structures in the Po plain by focusing specifically on the portion of the fault

segments that are located in Mesozoic carbonate sequences.

In summary, our results illustrate how a preliminary yet accurate image of the seismogenic

systems at depth may increase the robustness of seismotectonic models, especially in regions where

active faults are subdued or not at all exposed at surface.

Acknowledgments

This study has benefited from funding provided by the Italian Presidenza del Consiglio dei Ministri

– Dipartimento della Protezione Civile (DPC). This paper does not necessarily represent DPC

official opinion and policies. The authors thank the editor L. Jolivet, an anonymous reviewer and R.

Caputo for helpful reviews that strengthened this paper.

References

Ahmad, M.I., Dubey, A.K., Toscani, G., Bonini, L., Seno, S., 2014. Kinematic evolution of thrusts

wedge and erratic line length balancing: insights from deformed sandbox models.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

19

International Journal of Earth Sciences (Geol. Rundsch.) 133, 329-347, doi 10.1007/s00531-

013-0947-8.

Argnani, A., Ricci Lucchi, F., 2001. Terziary silicoclastic turbidite systems of the Northern

Apennines. In: Vai, G. B. and I. P. Martini (Eds.), Anatomy of an Orogen: the Apennines and

adjacent mediterranean basins, Kluwer Academic Publishers, 327-350.

Bertotti, G., Mosca, P., 2009. Late-orogenic vertical movements within the arc of the SW Alps and

Ligurian Alps. Tectonophysics 475(1), 117-127.

Bertotti, G., Capozzi, R., Picotti, V., 1997. Extension controls Quaternary tectonics,

geomorphology and sedimentations of the N-Apennines foothills and adjacent Po Plain

(Italy). Tectonophysics 282, 291-301.

Bertotti, G., Picotti, V., Bernoulli, D., Castellarin, A., 1993. From rifting to drifting: tectonic

evolution of the Southalpine upper crust from the Triassic to the Early Cretaceous.

Sedimentary Geology 86, 53-76, doi: 10.1016/0037-0738(93)90133-P.

Bigi, G., Cosentino, D., Parotto, M., Sartori, R., Scandone, P., 1990. Structural Model of Italy and

Gravity Map, 1:500,000. Quad. Ric. Scientifica, 114, 3, S.EL.CA Florence.

Bignami, C., Burrato, P., Cannelli, V., Chini, M., Falcucci, E., Ferretti, A., Gori, S., Kyriakopoulos,

C., Melini, D., Moro, M., Novali, F., Saroli, M., Stramondo, S., Valensise, G., Vannoli, P.,

2012. Coseismic deformation pattern of the Emilia 2012 seismic sequence imaged by

Radarsat-1 interferometry. Annals of Geophysics, 55, 4, doi:10.4401/ag-6157.

Boccaletti, M., Corti, G., Martelli, L., 2011. Recent and active tectonics of the external zone of the

Northern Apennines (Italy). Int. J. Earth Sci. (Geol. Rundsch.) 100, 1331-1348,

doi:10.1007/s00531-010-0545-y.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

20

Boccaletti, M., Bonini, M., Corti, G., Gasperini, P., Martelli, L., Piccardi, L., Severi, P., Vannucci,

G., 2004. Seismotectonic map of the Emilia-Romagna Region, Emilia-Romagna Region-

SGSS and CNR-IGG, S.EL.CA, Florence.

Bonini, L., Di Bucci, D., Toscani, G., Seno, S., Valensise, G., 2014. On the complexity of surface

ruptures during normal faulting earthquakes: excerpts from the 6 April 2009, L’Aquila

(central Italy) earthquake (Mw 6.3). Solid Earth.

Burrato, P., Ciucci, F., Valensise, G., 2003. An inventory of river anomalies in the Po plain,

Northern Italy: Evidence for active blind thrust faulting. Annals of Geophysics 46, 865–882.

Carena, S., Suppe, J., 2002. Three-dimensional imaging of active structures using earthquake

aftershocks: the Northridge thrust, California. Journal of Structural Geology 24, 887-904.

Cassano, E., Anelli, L., Fichera, R., Cappelli, V., 1986. Pianura Padana. Interpretazione integrata di

dati geofisici e geologici, in: Proceedings of the 73° Meeting of the Società Geologica

Italiana, September 29-October 4, 1986, Rome, Italy, pp. 27.

Castellarin, A., Vai, G.B., 1986. Southalpine versus Po Plain Apenninic arcs. in: The origin of arcs,

Development in Geotectonic, Wezel, F.C. (Ed), 253-280, Elsevier Amsterdam.

Dahlstrom, C.D.A., 1969. Balanced cross sections. Canadian Journal of Earth Sciences 6, 743-757.

Devoti, R., Esposito, A., Pietrantonio, G., Pisani, A.R., Riguzzi, F., 2011. Evidence of large scale

deformation patterns from GPS data in the Italian subduction boundary. Earth and Planetary

Science Letters 311, 230-241.

Dewey, J.F., Helman, M.L., Turco, E., Hutton, D.H.W., Knot, S.D., 1989. Kinematics of the

Western Mediterranean. In Coward M.P., Dietrich D. Ed.s: Alpine Tectonics. Geol. Soc.

London Spec. Publ. 45, 265-283.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

21

Di Domenica, A., Bonini, L., Calamita, F., Toscani, G., Galuppo, C., Seno, S., 2014. Analogue

modeling of positive inversion tectonics along differently oriented prethrusting normal faults:

an application to the Central-Northern Apennines of Italy. The Geological Society of America

Bulletin, doi: 10.1130/B31001.1.

DISS Working Group, 2010. Database of Individual Seismogenic Sources (DISS), Version 3.1.01:

A compilation of potential sources for earthquakes larger than M 5.5 in Italy and surrounding

areas, http://diss.rm.ingv.it/diss/.

Doglioni, C., 1991. A proposal for the kinematic modelling of W-dipping subductions - possible

applications to the Tyrrhenian-Apennines system. Terranova 3, 423-434.

Dondi, L., D’Andrea, G. 1986. La Pianura Padana e Veneta dall’Oligocene superiore al Pleistocene.

Giornale di Geologia 48, 197-225.

Elliott, D., 1982. The construction of balanced cross-sections. Journal of Structural Geology 5 ,

617–620.

Fantoni, R., Franciosi, R., 2010. Tectono-sedimentary setting of the Po Plain and Adriatic foreland.

Rend. Fis. Acc. Lincei 21, S197–S209, doi 10.1007/s12210-010-0102-4.

Fuis, G.S., Clayton, R.W., Davis, P.M., Ryberg, T., Lutter, W.J., Okaya, D.A., Hauksson, E.,

Prodehl, C., Murphy, J.M., Benthien, M.L., Baher, S.A.,. Kohler, M.D, Thygesen, K., Simila,

G., Keller G.R., 2003. Fault systems of the 1971 San Fernando and 1994 Northridge

earthquakes, southern California: relocated aftershocks and seismic images from LARSE II.

Geology 31, 171–174.

Ganas, A., Roumelioti, Z., Chousianitis, K., 2012. Static stress transfer from the May 20, 2012, M

6.1 Emilia-Romagna (northern Italy) earthquake using a co-seismic slip distribution model.

Annals of Geophysics 55, doi: 10.4401/ag-6176.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

22

Ghielmi, M., Minervini, M., Nini, C., Rogledi, S., Rossi, M., 2013. Late Miocene–Middle

Pleistocene sequences in the Po Plain – Northern Adriatic Sea (Italy): The stratigraphic record

of modification phases affecting a complex foreland basin. Marine and Petroleum Geology

42, 50-81.

Guidoboni E., Ferrari G., Mariotti D., Comastri A., Tarabusi G. and Valensise G., 2007.

CFTI4Med, Catalogue of Strong Earthquakes in Italy (461 B.C.-1997) and Mediterranean

Area (760 B.C.-1500). INGV-SGA. http://storing.ingv.it/cfti4med/

Harris, R.A., Simpson, R.W., Reasenberg, P.A. 1995. Influence of static stress changes on

earthquake locations in southern California. Nature, 375, 221– 224.

Haskell, N.A., 1964. Total energy spectral density of elastic wave radiation from propagating faults.

Bull. Seismol. Soc. Am. 54, 1811-1841.

Lin, J., Stein, R.S., 2004. Stress triggering in thrust and subduction earthquakes, and stress

interaction between the southern San Andreas and nearby thrust and strike-slip faults. Journal

of Geophysical Research 109, B02303, doi:10.1029/2003JB002607.

Maino, M., Decarlis, A., Felletti, F., Seno, S., 2013. Tectono-sedimentary evolution of the Tertiary

Piedmont Basin (NW Italy) within the Oligo–Miocene central Mediterranean geodynamics.

Tectonics 32, 593-619.

Malagnini, L., Herrmann, R.B., Munafo’, I., Buttinelli, M., Anselmi, M., Akinci, A., Boschi, E.,

2012. The 2012 Ferrara seismic sequence: Regional crustal structure, earthquake sources, and

seismic hazard. Geophys. Res. Lett. 39, doi: 10.1029/2012GL053214.

Malinverno, A., Ryan, W.B.F., 1986. Extension in the Tyrrhenian Sea and shortening in the

Apennines as a result of arc migration driven by sinking of the lithosphere. Tectonics 5, 227–

245.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

23

Mariotti, G., Doglioni, C., 2000. The dip of the foreland monocline in the Alps and Apennines.

Earth and Planetary Science Letters 181, 191-202.

Marone, C., Richardson, E., 2010. Learning to read fault-slip behaviour from fault-zone structure.

Geology 38, 767–768.

Marzorati, S., Carannante, S., Cattaneo, M., D’Alema, E., Frapiccini, M., Ladina, C., Monachesi,

G., Spallarossa, D., 2012. Automated control procedures and first results from the temporary

seismic monitoring of the 2012 Emilia sequence. Annals of Geophysics 55, doi: 10.4401/ag-

6116.

Montone, P., Mariucci, M.T., 1999. Active stress along the NE external margin of the Apennines:

the Ferrara arc, northern Italy. Journal of Geodynamics 28, 251-265,

http://dx.doi.org/10.1016/S0264-3707(98)00041-6.

Muttoni, G., Garzanti, E., Alfonsi, L., Cirilli, S., Germani, D., Lowrie, W., 2001. Motion of Africa

and Adria since the Permian: paleomagnetic and paleoclimatic constraints from northern

Libya. Earth Planet. Sci. Lett. 192, 159-174.

Nakajima, J., Hasegawa, A., Saeko Kita, S., 2011. Seismic evidence for reactivation of a buried

hydrated fault in the Pacific slab by the 2011 M9.0 Tohoku earthquake. Geophysical Research

Letters 38, L00G06, doi:10.1029/2011GL048432.

Nardon, S., Marzorati, D., Bernasconi, A., Cornini, S., Gonfalini, M., Mosconi, S., Romano, A.,

Terdich, P., 1990. Cavone oil field (Italy): an example of the application of integrated

methodologies to reservoir characterization. Memorie della Società Geologica Italiana 45,

791-805.

Noda, H., Lapusta, N., 2013. Stable creeping fault segments can become destructive as a result of

dynamic weakening. Nature, doi:10.1038/nature11703.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

24

Patacca, E., Scandone, P., Bellatalla, M., Perilli, N., Santini, U., 1992. La zona di giunzione tra

l’arco appenninico settentrionale e l’arco appenninico meridionale nell’Abruzzo e nel Molise.

Studi Geologici Camerti Special Volume: 1991–1992, 417-441.

Pezzo, G., Boncori, J.P.M., Tolomei, C., Salvi, S., Atzori, S., Antonili, A., Trasati, E., Novali, F.,

Serpelloni, E., Candela, L., Giuliani, R., 2013. Coseismic Deformation and Source Modeling

of the May 2012 Emilia (Northern Italy) Earthquakes. Seismological Research Letters 84, doi:

10.1785/0220120171.

Picotti, V., Pazzaglia, F.J., 2008. A new active tectonic model for the construction of the Northern

Apennines mountain front near Bologna (Italy). J. Geophys. Res. 113, B08412,

doi:10.1029/2007JB005307.

Pieri, M., Groppi, G., 1981. Subsurface geological structure of the Po Plain, Italy, in Progetto

Finalizzato Geodinamica, edited by C.N.R., 414.

Poliakov, A.N.B., Dmowska, R., Rice, J.R., 2002. Dynamic shear rupture interactions with fault

bends and off-axis secondary faulting. Journal of Geophysical Research 107, doi:

10.1029/2001JB000572.

Pondrelli, S., Salimbeni, S., Ekstrom, G., Morelli, A., Gasperini, P., Vannucci, G., 2006. The Italian

CMT dataset from 1977 to the present. Phys. Earth Planet. Int. 159, 286-303,

doi:10.1016/j.pepi.2006.07.008.

Pondrelli, S., Salimbeni, S., Perfetti, P., Danecek, P., 2012. Quick regional centroid moment tensor

solutions for the Emilia 2012 (northern Italy) seismic sequence. Annals of Geophysics 55,

doi: 10.4401/ag-6146.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

25

Ravaglia, A., Seno, S., Toscani, G., Fantoni, R., 2006. Mesozoic extension controlling the Southern

Alps thrust geometry under the Po Plain, Italy: insights from sandbox models. Journal of

Structural Geology 28, 2084-2096, doi: 10.1016/j.jsg.2006.07.011.

Ravaglia, A., Turrini, C., Seno, S., 2004. Mechanical stratigraphy as a factor controlling the

development of a sandbox transfer zone: A three-dimensional analysis. Journal of Structural

Geology 26, 2269-2283.

R.E.R. and ENI-AGIP, 1998. Riserve idriche sotterranee nella Regione Emilia-Romagna. G. Di Dio

eds., 119 pp., 9 sheets, S.EL.CA., Firenze.

Robertson, A.H.F., Grasso, M., 1995. Overview of the Late Tertiary Recent tectonic and palaeo-

environmental development of the Mediterranean region. Terra Nova 7, 114-127.

Roering, J., Cooke, M., Pollard, D., 1997. Why blind thrust faults do not propagate to the Earth's

surface: Numerical modeling of coseismic deformation associated with thrust-related

anticlines. Journal of Geophysical Research 102, doi: 10.1029/97JB00680.

Rogledi, S., 2010. Assetto strutturale delle unità alpine nella pianura tra il lago d’Iseo e il Garda.

Rischio sismico nella Pianura Padana, http://cesia.ing.unibs.it/index.php/it/eventi/giornate-di-

studio/119.

Royden L., Patacca, E., Scandone, P., 1987. Segmentation and configuration of subducted

lithosphere in Italy: an important control on thrust-belt and foredeep-basin evolution. Geology

15, 714-717.

Rovida, A., Camassi, R., Gasperini, P., Stucchi, M., 2011. CPTI11, the 2011 version of the

Parametric Catalogue of Italian Earthquakes, Milano, Bologna.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

26

Sarao, A., Peruzza, L., 2012. Fault-plane solutions from moment-tensor inversion and preliminary

Coulomb stress analysis for the Emilia Plain. Annals of Geophysics 55, 4, doi: 10.4401/ag-

6134.

Scholz, C.H., 1998. Earthquakes and friction laws. Nature , 391, 37-42.

Scisciani, V. 2009. Styles of positive inversion tectonics in the Central Apennines and in the

Adriatic foreland: Implications for the evolution of the Apennine chain (Italy). Journal of

Structural Geology 31, 1276-1294.

Scognamiglio, L., Margheriti, L., Mele, F.M., Tinti, E., Bono, A., De Gori, P., Lauciani, V.,

Lucente, F.P., Mandiello, A.G., Marcocci, C., Mazza, S., Pintore, S., Quintiliani, M., 2012.

The 2012 Pianura Padana Emiliana seimic sequence: locations and magnitudes. Annals of

Geophysics, 55, 549-559, doi:10.4401/ag-6159.

Scrocca, D., Carminati, E., Doglioni, C., Marcantoni, D., 2007. Slab retreat and active shortening

along the central-northern Apennines, in Thrust belts and Foreland Basins: from fold

kinematics to hydrocarbon systems. edited by O. Lacombe, J. Lavé, F. Roure and J. Verges,

471–487, Springer.

Sibson, R.H., 1985. Stopping earthquake ruptures at dilatational fault jogs. Nature 316, 248-251.

Stein, R.S., King, G.C.P., Lin, J., 1994. Stress triggering of the 1994 M = 6.7 Northridge,

California, earthquake by its predecessors. Science, 265, 1432–1435.

Tizzani, P., Castaldo, R., Solaro, G., Pepe, S., Bonano, M., Casu, F., Mamunta, M., Manzo, M.,

Pepe, A., Samsonov, S., Lanari, R., Sansosti, E., 2013. New insights into the 2012 Emilia

(Italy) seismic sequence through advanced numerical modeling of ground deformation InSAR

measurements. Geophysical Research Letters 40, 1971-1077.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

27

Toda, S., Stein, R.S., Richards-Dinger, K., Bozkurt, S., 2005. Forecasting the evolution of

seismicity in southern California: Animations built on earthquake stress transfer. Journal of

Geophysical Research 110, B05S16, doi:10.1029/2004JB003415.

Toscani, G., Burrato, P., Di Bucci, D., Seno, S., Valensise, G., 2009. Plio-Quaternary tectonic

evolution of the northern Apennines thrust fronts (Bologna-Ferrara section, Italy):

seismotectonic implications. Boll. Soc. Geol. Ital. (Italian Journal of Geosciences) 128, 605-

613, doi:10.3301.IJG2009.128.2.605.

Toscani, G., Seno, S., Fantoni, R., Rogledi, S., 2006. Geometry and timing of deformation inside a

structural arc: the case of the western Emilian folds (Northern Apennine front, Italy). Boll.

Soc. Geol. It. 125, 59-65.

Vanossi, M., Perotti, C.R., Seno, S., 1994. The Maritime Alps arc in the Ligurian and Tyrrhenian

systems. Tectonophysics 230 (1-2) , 75-89.

Valensise, G., Pantosti, D., 2001. The investigation of potential earthquake sources in peninsular

Italy: a review. Journal of Seismology 5 , 287-306.

Wang, W.H., Chen, C.H., 2001. Static stress transferred by the 1999 Chi-Chi, Taiwan, earthquake:

Effects on the stability of the surrounding fault systems and aftershock triggering with a 3D

fault-slip model. Bull. Seismol. Soc. Am., 91, 1041– 1052.

Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture

length, rupture width, rupture area, and surface displacement. Bulletin of the Seismological

Society of America 84, 974–1002.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

28

Figure captions

Figure 1. Map view of the Emilia area and structural map of the Ferrara arc showing the location

(http://iside.rm.ingv.it) and focal mechanism (Saraò and Peruzza, 2012) of the largest events

(M˃5) of the 2012 sequence. White dots show the location of M<5 events

(http://iside.rm.ingv.it). Black lines outline the buried Apennines thrust faults (from Bigi et

al., 1990). Dashed lines are the main buried inherited extensional faults mapped in the area

(Rogledi, 2010). A Mesozoic carbonate platform (Bagnolo platform) and Mesozoic

extensional basin (Suzzara basin) are also shown. The inset shows the location of the main

buried thrusts in the Po plain. S-S’ indicates the trace of the section shown in Figure 2.

Figure 2. a) Interpreted seismic line BO-365-88 (available at:

http://unmig.sviluppoeconomico.gov.it/videpi/en/), that cuts through the epicentral area of the

Emilia sequence (see Figure 1 for location). Velocity interval data used for depth conversion

are shown on the left. The inset show an example of extensional faults detected in the seismic

profile. b) Final geological interpretation of the seismic reflection profile (additional

stratigraphic units have been added based on literature data, e.g. Toscani et al., 2009), and

restored cross-section. The reference marker is the Late Eocene horizon that marks the change

from extension to compression in the Ferrara sector. The restoration of the Lower Pliocene-

Quaternary units has been omitted due to the complexity of syn-compressional erosion-

sedimentation processes.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

29

Figure 3. a) Geological cross-section with relocated events from Marzorati et al. (2012). Red stars

mark the location of the two mainshocks. Orange and yellow dots show the aftershocks of the

20 and 29 May sequences, respectively. b) Projection on the S-S’ cross section of the fault

planes used by different authors to model coseismic surface displacements. The upper panel

shows the maximum elevation change measured by InSAR analyses (Radarsat-2 data). c)

Histogram of the number of daily earthquakes and cumulative curve of events recorded in the

Emilia area by the INGV seismological network (http://terremoti.ingv.it/it/ultimi-eventi/842-

terremoti-in-pianura-padana-emiliana.html). Temporal relations between SAR image pairs

and the evolution of the sequence are also indicated.

Figure 4. Simplified down-dip geological section along the San martino and Mirandola thrust

showing the mainshock location (red star) and the coseismic slip distribution based on

seismological (a; Ganas et al., 2012), and geodetic data (b, and c; Pezzo et al., 2013).

Figure 5. a) Map view of the epicentral area of the Emilia sequence. SS’ represent the trace of the

geological sections. Broken lines are the section across the midpoint of the seismogenic

sources used to calculate stresses imparted by two mainshocks. The lower panel summarizes

the geometrical fault parameters used for modeling Coulomb stress changes.

Figure 6. a) Cross section of Coulomb stress changes associated with the 20 May (a) and 29 May

(b) earthquakes. White dots represents aftershocks located near the middle of the modeled

faults (thick red lines). The two upper panels in (a) and (b) show the stress calculated at the

middle of the modeled faults without seismicity (on the left) and with aftershocks (on the

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

30

right). Hypocentral location and focal mechanisms are also indicated. The lower panels in (a)

and (b) show stress transfer analyses plotted on the geological section.

Figure 7. a) Structural map of the top of Mesozoic carbonates (bottom of the Marne del Cerro

Formation; modified from Nardon et al., 1990), and (b) down-dip geological section of the

Mirandola structures. Red stars mark the M˃5 events of the 29 May sequence. In (b) the

coseismic slip pattern, (dotted lines, Pezzo et al., 2013) and the location of inherited faults

extrapolated from the structural map (a) are shown.

Figure 8. Range of fault slip behavior of the causative faults of the Emilia sequence. Red lines

represent the seismogenic rupture zones. Blue lines are fault segments where aseismic slip is

expected. Green lines highlight a hypothetical creeping fault segment. Yellow segments

indicate inherited extensional faults that may shown a partial reactivation during the sequence.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

31

Figure 1

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

32

Figure 2

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

33

Figure 3

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

34

Figure 4

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

35

Figure 5

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

36

Figure 6

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

37

Figure 7

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

38

Figure 8

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

39

Graphical Abstract

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

40

Highlights

Seismic reflection data allow us to image the active faults in the Emilia area

Seismological, geodetic and geological data are used to analyze the 2012 sequence

Active faults in Emilia show different rupture behaviour and downdip segmentation