Content uploaded by Paolo Boncio
Author content
All content in this area was uploaded by Paolo Boncio
Content may be subject to copyright.
Zoning Surface Rupture Hazard along Normal Faults: Insight
from the 2009 Mw6.3 L’Aquila, Central Italy, Earthquake
and Other Global Earthquakes
by Paolo Boncio, Paolo Galli,*Giuseppe Naso, and Alberto Pizzi
Abstract Following the observation of surface faulting in populated areas affected
by the 2009 Mw6.3 L’Aquila earthquake (central Apennines of Italy), we propose gen-
eral criteria for delineating zones of surface fault-rupture hazard (SFRH) along active
normal faults. Similar to other countries with surface faulting hazard, Italy does not
have explicit and comprehensive codes and/or regulations concerning this important
issue. Our proposal, which is explicitly inspired by the California Alquist–Priolo Earth-
quake Fault Zoning Act (A–P), compares the 2009 coseismic surface faults to surface
rupture data collected globally for several normal-faulting earthquakes. We propose
earthquake fault zones and fault setbacks, which are asymmetrically shaped around
the fault trace. The zones are wider on the hanging wall, consistent with the observation
of wider coseismic rupture zones in the hanging wall block compared to the footwall
block. For faults mapped in detail, we suggest a 150-m-wide earthquake fault zone
(EFZ) on the hanging wall and a 30-m-wide EFZ on the footwall. The suggested widths
of the fault setback (S) on the hanging wall and footwall are 40 m and 15 m, respectively.
Considering the data collected for the L’Aquila fault system and abroad, we are con-
fident that our proposal is conservative enough for Apennine-like normal faults.
Introduction
Surface fault-rupture hazard (SFRH) can be considered a
localized seismic hazard due to the breaching of the ground
surface from slip along a fault during a large earthquake.
This motion may offset, tilt, distort, and damage buildings on
or in the vicinity of the fault trace. SFRH also encompasses
fault creep, but this concerns only a few faults worldwide,
such as the creeping section of the San Andreas fault in cen-
tral California. Although SFRH should be one of the most
easily detectable seismic hazards, due to the visibility of ac-
tive fault traces, the 6 April 2009 Mw6.3 L’Aquila earth-
quake in central Italy demonstrates that there is much
progress to be made in assessing the hazard even in devel-
oped countries exposed to seismic risk. Indeed, the 2009 nor-
mal-faulting surface ruptures occurred across populated
areas producing mild-to-moderate damage to infrastructure
(e.g., pipelines, roads) and buildings, including structures
less than a few years old (EMERGEO Working Group,
2009;Falcucci et al., 2009;Galli et al., 2009;Boncio et al.,
2010;Galli et al., 2010;Vittori et al., 2011).
A valuable reference point aimed at mitigating SFRH is
the Alquist–Priolo Earthquake Fault Zoning Act (A–P),
which was adopted by the state of California (U.S.A.) in
1972 (Bryan and Hart, 2007). The A–P defines an earth-
quake fault zone (EFZ) as a regulatory zone around the sur-
face traces of active faults within which fault-rupture hazard
may occur and in which detailed investigations (geological,
high-resolution geophysical, paleoseismological) are re-
quired prior to building structures for human occupancy.
The A–P also defines a fault setback (S) as the distance from
the active fault trace within which critical facilities and struc-
tures designed for human occupancy cannot be built. The
minimum setback from an active fault is 50 feet (15 m) un-
less proven otherwise.
In Italy, there are no comparable regulations for new or
retrofit construction on or near active faults. The national
regulations for building design in seismically active areas
do not consider SFRH (NTC, 2008), and the only official
Italian publication addressing this problem is the guidelines
and criteria for seismic microzonation (Working Group MS,
2008). These guidelines define specific microzones, referred
to as zones of instability, surrounding active faults. Even if
not explicitly stated, the Working Group MS (2008) was in-
spired by the A–P. The Working Group MS (2008) refers to a
case study on the Norcia fault in central Italy (Galli et al.,
2005) and suggests the use of zones centered on the fault
*Also at CNR - Istituto di Geologia Ambientale e Geoingegneria,
Montelibretti, Italy.
918
Bulletin of the Seismological Society of America, Vol. 102, No. 3, pp. 918–935, June 2012, doi: 10.1785/0120100301
trace, with setbacks of 15 m for known fault traces and
75 m for uncertain fault traces. However, there is not an ex-
plicit distinction between EFZs and Ss, and the general cri-
teria for defining the shape and width of the zones are not
explained.
The Eurocode 8 (2003; part 5, section 4.1.2: Proximity
to seismically active faults) states that buildings important
for public safety shall not be erected in the immediate vici-
nity of seismically active faults. A fault is considered active if
it shows evidence of movement during the late Quaternary.
Moreover, special geological investigations shall be carried
out for urban planning and for the erection of important
structures near potentially active faults in areas of high seis-
micity. However, there are no specifics on what immediate
vicinity means in terms of distance from the trace of an active
fault (e.g., the setback) or on how wide the zone around a
potentially active fault (e.g., the EFZ) should be for conduct-
ing special geological investigations.
In this paper, we summarize the main geologic features
of the surface ruptures that occurred during the 2009 Mw6.3
L’Aquila earthquake with particular attention given to the
populated Paganica and San Gregorio areas. The character-
istics of the 2009 surface ruptures are compared with data
collected from the literature for several other normal-faulting
earthquake surface ruptures worldwide. Our main purpose is
(1) to provide insight on how field observations can be trans-
lated into general criteria for shaping zones of local SFRH
along Apennine-type normal faults, applicable to Italy and
other areas with comparable seismotectonic settings and seis-
mic hazard, and (2) to provide a case study to help define
basic criteria for establishing a minimum width for EFZs
and Ss along active normal faults, even for regions that
already have regulations for SFRH.
Seismotectonic Setting
The 6 April 2009 Mw6.3 L’Aquila earthquake is a
normal-faulting earthquake that reactivated the Mount
Stabiata–Paganica–San Demetrio (SPD) fault system, a
southwest-dipping normal fault system (Fig. 1;SPD system
in Boncio et al., 2010; Paganica–San Demetrio fault system
in Galli et al., 2010). The mainshock (1:32 UTC) nucleated at
a depth of ∼9:5km (Chiarabba et al., 2009;Pondrelli et al.,
2010), close to the northwestern termination of the fault
Figure 1. (a) Map of the 2009 L’Aquila earthquake epicentral area with traces of the SPD normal fault system. (b) Five-meter-resolution
digital elevation model of the Paganica–San Gregorio populated areas showing the locations of active faults and traces of the 2009 coseismic
ground ruptures.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 919
system. The rupture propagated to the southeast with a sub-
surface length of 18–20 km (Anzidei et al., 2009;Atzori
et al., 2009;Cirella et al., 2009;Walters et al., 2009). An
Mw5.6 aftershock occurred on 7 April about 9 km southeast
of the mainshock at a depth of 15 km, within the SPD system.
Overall, the main seismogenic source strikes 130°–135° and
dips southwest at 50°–55°. A second strong aftershock oc-
curred on 9 April (Mw5.4, 10–11 km depth) about 18 km
north of the mainshock on the southwest-dipping normal
Laga Mountains fault (Blumetti et al., 1993;Galadini and
Galli, 2000;Boncio, Lavecchia, et al., 2004). The fault
parameters are similar to the 6 April mainshock (strike 130°,
dip 50° southwest, nearly dip-slip; Chiarabba et al., 2009;
Pondrelli et al., 2010).
The 6 April earthquake occurred just beneath L’Aquila
(population ∼73;000), which suffered severe damage (VIII–
IX on the Mercalli–Cancani–Sieberg [MCS] scale). Overall,
the earthquake produced damage up to IX–XMCS, with a
damage area of >VII–VIII MCS extending ∼22 km in the
northwest–southeast direction along the Aterno River valley,
mostly on the hanging wall of the SPD system (Galli et al.,
2009). The impact on the local inhabitants was enormous,
with 308 deaths and more than 67,000 left homeless.
The SPD and Laga Mountains faults belong to a regio-
nal-scale system of active normal faults that extends for hun-
dreds of kilometers along the axial part of the Italian
Apennines, from northern Tuscany to Calabria (Calamita and
Pizzi, 1994;Lavecchia et al., 1994;Boncio et al., 2000;Ga-
ladini and Galli, 2000;Pizzi et al., 2002;Boncio, Lavecchia,
and Pace, 2004;Galli et al., 2008;DISS Working Group,
2010 and references therein). The active Apennine normal
faults have the following common features: (a) they postdate
previous compressional structures of the Tertiary Apennine
orogeny; (b) they control the formation and evolution of
Quaternary or Pliocene–Quaternary intramontane basins;
(c) they have prevailing dip-slip kinematics; (d) they are
formed by long isolated segments or systems of short linked
segments; (e) they show evidence of repeated motion during
the late Quaternary; (f) the footwall blocks are composed of
Meso–Cenozoic rocks, mostly carbonates but also turbidites
and crystalline rocks; (g) they are characterized by fault
scarps, the evidence of which depends on slip rate, lithology,
and erosion rates; and (h) the footwall rocks along the fault
scarp are often in contact with slope-derived debris and col-
luvial deposits that accumulated in the hanging wall block
due to a combination of tectonic and morpho-sedimentary
events during the Quaternary glaciations. Most of the strong
Italian historical earthquakes (e.g., Working Group CPTI,
2004) can be associated with these active normal faults (Galli
et al., 2008).
The SPD system dips to the southwest, offsetting the
Meso–Cenozoic carbonates of the Gran Sasso range and bor-
dering the Aterno continental basin to the northeast (Bosi and
Bertini, 1970;Bagnaia et al., 1992;Bertini and Bosi, 1993;
Vezzani and Ghisetti, 1998;APAT, 2005). On the hanging
wall, Pleistocene to Holocene alluvial and colluvial deposits
are offset and lowered toward the Aterno basin (Galli et al.,
2010;Boncio et al., 2011).
During the 6 April mainshock, coseismic surface faulting
occurred along the SPD system. The ground ruptures were
mapped continuously along the Paganica fault segment and
found to have a maximum vertical offset of 10–15 cm at Pa-
ganica village (Boncio et al., 2010;EMERGEO Working
Group, 2009;Falcucci et al., 2009;Galli et al., 2009;2010;
Vittori et al., 2011). The estimated total length of the coseis-
mic surface faulting and fracturing ranges from ∼10 km
(Falcucci et al., 2009)to∼13 km (Boncio et al., 2010)to
∼20 km (Galli et al., 2009;Galli et al., 2010) in the north-
west–southeast direction along the SPD system. More conser-
vative values are reported by the EMERGEO Working Group
(2009) (2.5 km up to possibly 6 km) and Vittori et al. (2011)
(≥2:6km). Vittori et al. (2011) infer that the ruptures ob-
served by Boncio et al. (2010) and Galli et al. (2009) north-
west of Paganica (up to Mount Stabiata) and south-southeast
of Paganica (San Gregorio) are sympathetic slip on secondary
faults.
Coseismic Ground Ruptures across the Populated
Paganica and San Gregorio Areas
The field survey performed along the Paganica and San
Gregorio normal faults allowed us to observe a number of
ground ruptures in the form of (a) en échelon or linear frac-
tures with centimeter-size apertures and normal dip-slip
motion of a few centimeters down to the southwest, (b) linear
fissures with centimeter-size openings without vertical slip,
and (c) en échelon cracks. In general, the apertures and dip-
slip motions of the fractures range from a few millimeters to
a maximum of 10–15 cm, with most between 1 and 6 cm
(Boncio et al., 2010). The ruptures indistinctly cut the
ground surface, roads, and man-made structures of various
materials, including reinforced concrete.
We grouped the observed ground deformation into four
main types according to the type of deformation observed at
the surface and its relation with the seismogenic fault.
•Type 1 deformation can be considered coseismic surface
faulting in a strict sense. The seismic rupture propagated
upward along the main fault, breaching the ground surface
coseismically. This deformation appears as free faces sev-
eral centimeters tall, with centimeter-size apertures, at the
base of well-preserved preexisting fault scarps in strongly
lithified rocks (e.g., conglomerate, limestone; Fig. 2a,b).
Where this type of deformation crosses buildings, the
damage is localized along the fault trace, producing open
fractures with centimeter-size steps or fractures accompa-
nied by tilting of the buildings toward the hanging wall
(Fig. 2c–e).
•Type 2 is also coseismic surface faulting in a strict sense,
but the seismic slip breached the surface along synthetic
splays, cutting through poorly consolidated deposits
(e.g., colluvium or gravels) in the hanging wall of the main
920 P. Boncio, P. Galli, G. Naso, and A. Pizzi
fault (Fig. 3). This resulted in a more irregular distribution
of displacement including the following: (a) an offset of the
ground surface along a single fracture; (b) an offset distrib-
uted among parallel fractures spaced a few meters apart; or
(c) open fissures with a warped ground surface. Synthetic
splays occurred from a few meters to ∼35 m from the main
fault trace (Fig. 3), with some at larger distances (120–
140 m; Fig. 4). The damage associated with this type
of deformation is similar to that associated with Type 1
ruptures.
For both Types 1 and 2, the vertical displacement (d) is
systematically associated with the horizontal aperture (o)
of the fault and the ratio between the two components
(d/o) ranges from ∼1to ∼2:5(Boncio et al., 2010). This
suggests a nonplanarity of the fault, steepening upwards,
that allowed the fault walls to lose contact during slip near
the surface. For Type 1, the d/o ratio ranges from 1.2 to 2.5.
The dip of the fault at the surface is constrained by field
measurements to 60° to 80°, and this suggests a dip in the
subsurface ranging from 40° to 50°, in agreement with seis-
mological data (average dip of 50°–55°; Chiarabba et al.,
2009;Pondrelli et al., 2010).
•Type 3 deformation is characterized by open fissures along
the trace of the main fault, where the fault is covered by
unconsolidated colluvium or alluvium (Fig. 5). These fis-
sures are accompanied by flexure of the hanging wall with
a wavelength of a few meters. This flexure was amplified
during the early postseismic phase. In several places, a fis-
sure along the fault trace was observed shortly after the
mainshock. Several days after the mainshock, the hanging
wall flexure showed an appreciable vertical displacement
of a few centimeters (Fig. 5b). This amplification of the
hanging wall flexure, and of the overall displacement, can
be reasonably explained as related to afterslip processes. In
general, Type 3 deformation can be interpreted as the su-
perficial expression of blind normal faulting with the upper
tip of coseismic faulting perhaps located at the interface
between the stiff bedrock and the soft colluvial–alluvial
Figure 2. Examples of Type 1 coseismic deformation (breaching of the ground surface during coseismic slip along the main fault).
(a) Field view of the main Paganica fault. The fault scarp is well preserved in conglomerates on the footwall block (detailed location
in Fig. 4). (b) Detail of (a) showing the 2009 reactivation at the base of the main fault scarp (f.f., free face, i.e., rejuvenated part of
the preexisting fault scarp; o, opening; t, throw). (c–e) Examples of damage from Type 1 coseismic faulting (detailed location in Fig. 3).
The color version of this figure is available only in the electronic edition.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 921
Figure 3. Orthophoto map of Paganica north (see Fig. 1) with examples of Type 2 coseismic deformation (breaching of the ground
surface during coseismic slip along a synthetic splay of the main fault). The color version of this figure is available only in the electronic
edition.
Figure 4. Orthophoto map of Paganica south (see Fig. 1) with examples of coseismic surface faulting along a synthetic splay of the main
fault (Type 2 coseismic deformation) located far from the main fault trace (120–140 m). The color version of this figure is available only in the
electronic edition.
922 P. Boncio, P. Galli, G. Naso, and A. Pizzi
covers. Based on observations and models for shallow
blind normal faults (Martel and Langley, 2006;Kaven
and Martel, 2007), we interpret the coseismic fissures as
being due to tensional stresses originating between the
tip of the coseismic fault and the ground surface. Dip-
slip motion at depth caused the hanging wall flexure,
and the increased dip-slip motion due to afterslip processes
amplified the surface flexure. In some places there is
evidence that the afterslip propagated upward along the
fracture, producing centimeter-size steps on the former fis-
sures. Considering the short wavelength of the hanging
wall flexure (e.g., 2.5–3 m in Fig. 5b), we infer a very shal-
low tip of the coseismic faulting, perhaps as shallow as a
few meters.
•Type 4 deformation is typical of the San Gregorio area and
is characterized by linear fissures and en échelon cracks
without appreciable vertical displacement, located on the
hanging wall of the seismogenic fault (Fig. 6). Electrical
resistivity tomography and geognostic investigations al-
lowed us to reasonably constrain the subsurface geometry
of the San Gregorio normal fault (Fig. 6b,c;Boncio et al.,
2011). By projecting the fault plane constrained by geo-
physical and geognostic investigations up-dip to the
ground surface, we measured a distance of 60 to 100 m
between the ground ruptures and the projected fault trace
(Fig. 6). Other fissures were observed on the footwall of
the San Gregorio normal fault close to the southeastern tip
of the coseismic deformation zone. Nevertheless, it is un-
clear if these fissures can be related to the San Gregorio
normal fault or to other minor structures reactivated during
the earthquake. The hanging wall fissures may have origi-
nated because of near-surface tensional stresses above the
tip of a blind coseismic normal fault (Martel and Langley,
2006;Kaven and Martel, 2007). Alternatively, the hanging
wall fissures may be interpreted as mode I cracks that pro-
pagated upward coseismically from the tip of the blind co-
seismic fault. For the latter case, considering the distance
of the fissures from the fault trace (60–100 m) and a dip of
60°–70° for the San Gregorio normal fault, and assuming
subvertical mode I cracks, we infer a depth of 100 to 280 m
for the tip of the coseismic fault.
Architecture of the Main Active Fault Zone from
Trench Observations
Coseismic faulting in the northern part of Paganica vil-
lage caused rupture of the Gran Sasso water pipe (see Fig. 3
for location). Water escaped at high pressure and excavated
a deep gorge orthogonal to the Paganica fault, exposing the
fault zone architecture (Fig. 7). The stratigraphic succession
was investigated by detailed mapping, with ages constrained
by 14C and U/Th dating and tephrochronology. The unit
succession spans from the Middle (Early?) Pleistocene to
the Holocene and includes breccias, alluvial gravels, paleo-
sols, and bodies of colluvial deposits (see Boncio et al.,
2010 and Galli et al., 2010 for further details).The fault
zone is 40–45 m wide and is formed by at least 5 steeply-
dipping synthetic normal faults (F1 to F5 in Fig. 7). With
the exception of F3 and F3′, which are very close to each
other and probably join at a shallow depth, the faults are
spaced 10–12 m apart. Fault F2 fits with the main Paganica
normal fault as mapped on a 1:5000 scale topographic map,
based on geologic and geomorphologic features (fault trace
in Fig. 3). The oldest sedimentary unit crops out at the foot-
wall of F2, in agreement with the interpretation that F2 is
the main fault of the entire fault zone. Therefore, F1 is a
footwall synthetic splay and F3 to F5 are hanging wall syn-
thetic splays of the main Paganica normal fault (F2). The
2009 coseismic faulting occurred on the F5 splay (Type
2 deformation). The 2009 fault throw was ∼10 cm, as in-
dicated by the offset of the bottom of the sedimentary unit
formed by historical colluvium and modern soil (U2 in
Fig. 7). A fracture was mapped at the surface along the
F5 fault trace.
A detailed paleoseismologic interpretation of the
trench indicates that F5 is not the only active strand of
the fault zone. Evidence of repeated offset during the Late
Pleistocene-Holocene was observed along the F3′, F4, and
Figure 5. Examples of Type 3 coseismic deformation (see Fig. 3for locations). (a) Open fissures along the trace of the main fault in
unconsolidated colluvium. (b) Open fissures along the trace of the main fault in unconsolidated alluvium, accompanied by flexure of the
hanging wall. The color version of this figure is available only in the electronic edition.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 923
F5 strands, with offsets much larger than that of 2009 (up to
80 cm; Falcucci et al., 2009;Boncio et al., 2010;Galli et al.,
2010;Cinti et al., 2011). About 70–80 m southeast of the
water pipe trench, the 2009 coseismic surface faulting oc-
curred along the trace of the main fault (Type 1 deformation).
Therefore, the entire fault zone exposed along the water pipe
trench must be considered an active fault zone, at least from
the main fault (F2) to the most distant splay at the hanging
wall (F5). This observation is important for constraining the
shape and size of the zone of SFRH.
Discussion
Width of the Rupture Zone (WRZ) for the L’Aquila
Earthquake: A Synthesis
Figure 8synthesizes the first-order structural character-
istics of the ground ruptures associated with the 6 April 2009
L’Aquila earthquake. Figure 8a shows the Paganica normal
fault, which was reactivated during the mainshock and rup-
tured to the ground surface. In addition to the evidence of
reactivation along the trace of the main fault (deformation
Figure 6. (a) Orthophoto map of the San Gregorio area with examples of Type 4 coseismic deformation (open fissures, without appreci-
able vertical displacement, in the hanging wall of the fault; o, opening). (b) Geologic section across the San Gregorio normal fault, con-
strained by electric resistivity tomography and well data, showing the relation between normal faulting and coseismic fissures. (c) Example of
shallow electric resistivity tomography showing the sharp resistivity contrast across the San Gregorio normal fault. The sections in (b) and (c)
are modified from Boncio et al. (2011). The color version of this figure is available only in the electronic edition.
924 P. Boncio, P. Galli, G. Naso, and A. Pizzi
Types 1 and 3), there is evidence of faulting and fracturing
along synthetic hanging wall splays, up to 30–35 m from the
trace of the main fault (Type 2 deformation). This evidence
of coseismic faulting and fracturing, together with the ob-
served architecture of the fault zone along the water pipe
trench, allows us to define a main 35-m-wide deformation
zone, delimited at the northeast by the trace of the main fault.
Reactivation of a synthetic splay located at larger distances
from the trace of the main fault (120–140 m) was also ob-
served. Nevertheless, this splay of short longitudinal conti-
nuity (∼170 m long) is not considered part of the main
deformation zone, but a subsidiary structure. The most
important difference between the main deformation zone
observed at the water pipe trench (m.d.z. in Fig. 8a) and the
aforementioned subsidiary structure (hanging wall splay in
Fig. 8a) is that while the main deformation zone can reason-
ably extend along strike for the entire length of the main
fault; the subsidiary hanging wall splays may be of much
shorter longitudinal continuity.
Figure 8b represents the case of the San Gregorio normal
fault. The slip along the fault did not reach the ground sur-
face and the deformation at the surface appears as open
fissures in the hanging wall, without appreciable vertical dis-
placement, parallel to the main fault and located 60–100 m
from the trace of the main fault.
The data from the Paganica and San Gregorio faults
might help in determining the shape and size of the EFZ
and S(e.g., Bryan and Hart, 2007,A–P).
WRZ Data from Normal-Faulting
Earthquakes Globally
We collected surface rupture data for several normal-
faulting earthquakes worldwide, ranging in magnitude from
5.6 to 7.6, in order to compare them with the observations
Figure 7. (a) Field view of the water pipe trench at Paganica north (location in Fig. 3). (b) Log of the northwestern wall of the trench
showing the architecture of the Paganica fault zone (slightly modified from Boncio et al., 2010). F2 correlates with the main Paganica fault as
mapped from geologic and morphologic field evidence (fault trace in Fig. 3), F1 is a footwall splay, and F2 to F5 define the main active
deformation zone. The 6 April 2009 coseismic faulting occurred on F5. The color version of this figure is available only in the electronic
edition.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 925
from the L’Aquila earthquake. The results are synthesized in
Table 1and displayed as histograms in Figure 9. The WRZ is
considered to be the width of the zone affected by coseismic
surface ruptures, measured perpendicular to the reactivated
main fault. In the case of broad deformation zones without
a well-defined main fault, such as those observed in major
geometric complexities separating adjacent faults (e.g.,
kilometer-scale stepovers), we measured the WRZ perpendi-
cular to the average strike of the deformation zone. We col-
lected the WRZ data reported in published papers, such as for
the 1987 Edgecumbe earthquake in New Zealand (Beanland
et al., 1989), or in most of the cases, we measured the WRZ
from published maps. The errors in measuring the WRZ
depend on the detail of the original survey and the scale of
the published maps. We measured WRZ only from detailed
maps (uncertain or inferred ruptures were not considered).
We are confident that the maximum errors of the measured
values are on the order of 10%. We considered dip-slip nor-
mal-faulting earthquakes, as well as a number of cases with
normal-oblique kinematics (e.g., 1954 M7.2 Fairview Peak,
1970 M7.1 Gediz, 1975 M5.9 Oroville; see Table 1).
The most common feature of the analyzed rupture zones
is a single rupture along a main preexisting range-front nor-
mal fault with a WRZ of a few meters or less (WRZ 1in
Table 1and Fig. 9). The reactivation of the main fault is com-
monly accompanied by secondary structures such as syn-
thetic and antithetic minor faults or cracks parallel to or
bifurcating from the main fault, with consequent wider rup-
ture zones. The secondary structures reactivate preexisting
fault scarps (i.e., features detectable prior to the earthquake)
or occur as new features (i.e., undetectable prior to the earth-
quake). En échelon ruptures are common, especially for ob-
lique-slip earthquakes or at geometric complexities of the
main fault for dip-slip earthquakes (e.g., lateral terminations
of the main rupture, kilometer-scale stepovers, and salients or
bends; see examples in Fig. 9c–f).
The coseismic ruptures are located along the main fault
and on the hanging wall. Footwall splays are not systematic
features in the analyzed earthquakes; they occur as local short
features, generally reactivating preexisting fault scarps lo-
cated at distances of 25 to 780 m from the main fault. Overall,
footwall splays are absent or less than 1.5%–2.5% of the entire
surface rupture length (SRL; see Table 1). An exception is the
1959 M7.3 Hebgen Lake earthquake, which had three foot-
wall splays at up to 780 m from the main fault, covering 6.5%
of SRL, and occurring near a major geometric complexity of
the main fault (a kilometer-scale fault bend; see Table 1).
Although WRZ data span from 1 to 1350 m, the WRZ is
generally less than 120–150 m for a simple fault trace
without internal major geometric or structural complexities
(Table 1, Fig. 9a,b). In general, a WRZ >150 m is asso-
ciated with kilometer-scale geometric complexities, such as
(a) stepovers between major en echelon segments (e.g., the
Bend along the 1954 Dixie Valley fault rupture, Fig. 9c;
the stepover between the Pearce and Sou Hills faults during
the 1915 Pleasant Valley earthquake), (b) salients (Bell Can-
yon salient along the 1954 Fairview Peak fault rupture,
Fig. 9d; West Spring block along the 1983 Lost River fault
rupture, Fig. 9f), (c) sharp bends (Red Canyon fault rupture
during the 1959 Hebgen earthquake), and (d) other structural
features such as a hanging wall graben (the Graben along the
1983 Lost River fault rupture).
Overall, more than 95% of the SRL along noncomplex
fault traces is characterized by ruptures within the hanging
wall at distances less than 150 m from the main fault, with a
significant percentage (>80%–90%) within 40 m of the
main fault. A significant divergence from these percentages
may characterize normal-oblique faults (e.g., Fairview Peak,
Gediz, Oroville). This suggests that along a 50°–60° dipping
fault (i.e., a typical normal fault), large strike-slip compo-
nents of motion may determine strain partitioning in the
hanging wall, with consequently wider rupture zones. There-
fore, both fault geometry and kinematics are important in
estimating maximum WRZs.
In Figure 10, we plotted the collected WRZ data (includ-
ing WRZs of geometrical complexities) against magnitude.
The WRZ seems to scale with magnitude. This positive rela-
tion between WRZ and magnitude is explained by the fact
that for large magnitude earthquakes the rupture involves
more than one simple fault segment; it crosses geometric
Figure 8. Sketches (not to scale) synthesizing the occurrence of
coseismic surface faulting/fracturing along the (a) Paganica and
(b) San Gregorio normal faults.
926 P. Boncio, P. Galli, G. Naso, and A. Pizzi
Table 1
Fault Parameters and WRZ for Normal-Faulting Earthquakes Worldwide
Earthquake;
Magnitude Fault Kin. *
SRL
(km)
MD †
(m) WRZ ‡(m) Ref. §Notes ∥
% SRL #
HW ≤150
% SRL #
HW ≤40
% SRL #
FW
1915 Avezzano,
Italy; 7.0
Fucino system N 36 1.0 (v)1–5 (1),
1–40 (2)
GG99 (1) Measured in paleoseismological trenches.
(2) Measured at the surface.
100 100 0
1915 Pleasant
Valley,
Nevada, U.S.A.;
7.6
China Mt., Tobin,
Pearce, Sou Hills
N 59 5.8 (v)1–195 (mostly
≤120) (1),
750–1350 (2)
Wa84 (1) SR along the main range-front faults. Three local FW
splays 115–670 m long, distant up to 50–135 m from
MF (Pearce fault). (2) Broad discontinuous zone of SR
parallel to MF (Pearce fault), 4.5 km long, in the
stepover zone between Pearce and Sou Hills faults
(separation between faults from 2.5 km to 5 km).
98.0
(calculated
for [1])
90.6 2.2
1946 Ancash,
Peru;
6.8
Quiches N 21 3.5 (v)1–70 Be91 WRZ is calculated for the 5.5-km-long Llamacorral
segment on a ∼1:43,500 tectonic map and a ∼1:3700
map (Fig. 2 in Be91). Local FW splay, 145 m long,
distant up to 25–30 m from MF (central part of the
segment).
100 94.2 2.6
1950 Fort Sage
Mts., California,
U.S.A.; 5.6
Fort Sage Mts. N 8.85 0.2 (v)1–380 Gi57 Large WRZ (280–380 m) due to a 360-m-long (4.1% of
SRL) antithetic HW splay.
95.9 95.9 0
1954 Fairview
Peak, Nevada,
U.S.A.; 7.2
Fairview Peak NR to RN 31.6 3.8 (v),
2.9 (h)
1–1010
(mostly ≤150)
Ca96 WRZ >150 m for 6.6 km (20.9% of SRL) along left-
stepping and right-stepping bifurcations from MF at the
Bell Canyon salient (major geometric complex-
ity of MF) and at the US Highway 50 left-stepping and
parallel segments. Two local FW splays, 170–270 m
long, distant up to 305 m from MF.
79.1 70.2 1.4
West Gate N to RN 10 1.1 (v),
1.2 (h)
1–85 Ca96 Two parallel SR, separated by 220–270 m, overlapping
for 395 m at the southern termination of the main
rupture zone. It is unclear which one is the MF
(possible FW splay, ?). Probably, partial reactivation of
the stepover zone between two major left-stepping
segments.
100 95.7 0 (?)
Louderback Mts. RN 14 0.8 (v),
1.7 (h)
1–120 Ca96 Discontinuous SR, mostly along left-stepping en echelon
segments. Overlapping zones between adjacent
segments (separation ranging from 140 to 195 m) are
not considered in calculating WRZ; WRZ is calculated
for each segment.
100 98.4 0
Gold King N 8.5 1.0 (v)1–40 Ca96 Discontinuous SR. Overlapping zones between en
echelon segments (separation ranging from 305 to
625 m) are not considered in calculating WRZ; WRZ
is calculated for each segment.
100 100 0
Phillips Wash LN 6.2 0.5 (v),
0.8 (h)
1–580 Ca96 Discontinuous complex SR along right-stepping en
echelon and/or parallel segments. Maximum WRZ
(580 m) for 1.4-km-long system of 3 parallel segments
(central part of SR zone).
56.4 53.6 0
(continued)
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 927
Table 1 (Continued)
Earthquake;
Magnitude Fault Kin. *
SRL
(km)
MD †
(m) WRZ ‡(m) Ref. §Notes ∥
% SRL #
HW ≤150
% SRL #
HW ≤40
% SRL #
FW
1954 Dixie Valley,
Nevada, U.S.A.;
6.8
Dixie Valley N 42 2.8 (v)1–400 (mostly
≤120)
(1), 1–705 (2)
Ca96 (1) SR along the main range-front fault. WRZ >120 m
only for 2.28 km (5.4% of SRL) along left-stepping
bifurcations from MF near the Bend (major geometric
complexity of MF; i.e., relay zone between northern
and southern Dixie Valley segments) and at the
southern termination of the MF. Two local FW splays
285–320 m long, distant up to 140–260 m from MF
near the Bend. (2) Broad discontinuous zone of SR on
the piedmont of the Bend area (i.e., major geometric
complexity of MF).
95.6
(calculated
for [1])
92.9 1.4
1959 Hebgen
Lake, Montana,
U.S.A.; 7.3
Hebgen, Red Canyon,
West Yellowstone
Basin
N 26.5 6.1 (v)1–300
(mostly <130)
Wi62 WRZ >150 m for 1.24 km (4.7% of SRL) along the
sharp bend between northern and southern segments of
the Red Canyon fault (major geometric complexity of
MF). Three local FW splays 440–800 m long, distant
90–780 m from MF (near the bend of Red Canyon
fault).
95.3 91.3 6.5
1970 Gediz,
Turkey; 7.1
Akcaalan, Pinarbasi,
Erdogmus, Sazkoy,
Muratdag
NL
to N
42 2.75 (v),
0.8 (h)
1–285
(mostly <130)
Ta71 WRZ >130 m in broad deformation zones at lateral
terminations of major fault segments (bifurcations
from MF, systems of en echelon fractures). One HW
splay, 175 m long, distant 170–210 m from MF
(Akcaalan segment). Local FW splays, 170–190 m
long, distant 30–55 m from MF (Akcaalan segment).
98.4 89.9 1.3
1975 Oroville,
California,
U.S.A.; 5.9
Cleveland Hill N to
NR
3.8 0.55 (v)1–450 (1),
1–30 (2)
Cl76 (1) Large WRZ (260–450 m) results from overlapping of
two right-stepping major fault segments (a western
segment, formed by the south and northwest colinear
breaks of Cl76, and an eastern segment, formed by the
northeast break of Cl76). (2) WRZ measured
individually for the two right-stepping segments. Local
short footwall crack, ∼60 m long, distant ∼30 mfrom
MF (south break).
35.5 (1),
100 (2)
35.5 (1),
100 (2)
1.6
1980 Irpinia, Italy;
6.9
Irpinia N 30 1.3 (v)1–25 PV90 Measured in paleoseismological trenches. 100 100 0
1981 Corinth,
Greece; 6.7, 6.4
(1); 6.4 (2) **
Pisia, Skinos (1);
Kaparelli (2)
N12–15 (1);
12.7 (2)
0.8 (v)
(1),
0.7 (v)
(2)
1–70 (1), 1–458
(mostly ≤115) (2)
Pa93,
PC04,
Ja82
(1) SR at or a few meters downslope of MF. Locally,
right-stepping en echelon fractures with WRZ of
∼70 m (Skinos, near the foot of an alluvial fan).
(2) WRZ <150 m along the main western and eastern
Kaparelli segments. Large WRZ (69–458 m) along a
broad discontinuous SR zone, 1.8 km long, between
the eastern and western Kaparelli segments (major
geometric complexity of MF, not considered in
calculating the percentages). Local short FW splay,
∼190 m long, distant 55–95 m from MF (western
Kaparelli segment).
100 89 (2),
not quanti-
fied for (1)
1.7 (2)
(continued)
928 P. Boncio, P. Galli, G. Naso, and A. Pizzi
Table 1 (Continued)
Earthquake;
Magnitude Fault Kin. *
SRL
(km)
MD †
(m) WRZ ‡(m) Ref. §Notes ∥
% SRL #
HW ≤150
% SRL #
HW ≤40
% SRL #
FW
1983 Borah Peak,
Idaho, U.S.A.;
7.3
Lost River N to
NL
33.3 2.7 (v),
1.0 (h)
1–780
(mostly
≤140)
Cr87 WRZ >140 m at major geometric complexities of MF
(e.g., 1.3-km-long West Spring Block, southern
section of MF, WRZ up to 780 m) and at an ∼1:7-km-
long HW graben, partially reactivated in 1983, up to
240 m wide (northern section of MF, Gooseberry
Creek). Local FW splay, 740 m long, distant up to
120 m from MF (southern section of MF, site E of
Cr87).
92.3 82.5 2.2
1986 Kalamata,
Greece; 5.8
Kalamata N 6 0.18 (v)1–60 Ly88 Detailed description lacking. Maximum WRZ obtained
from 1:59,000 tectonic map (Fig. 2b of Ly88).
100 ∼95 0
1987 Edgecumbe,
New Zealand;
6.3
Edgecumbe, Onepu,
Rotoitipakau
(preexisting), and
Awaiti, Otakiri,
Te Teko, Omeheu
(new)
N 21.7 (1);
16.3 (2)
2.5 (v)1–80 Be89 (1) Obtained by summing the length of each individual
fault. (2) Length of the system along the average strike.
WRZ exceeds 40 m only at 5 sites along the
Edgecumbe fault.
100 Not quantified
(>98)
0
1995 West
Macedonia;
6.6
Aliakmon River N 30 0.18 (v)1–70 Ch98,
Mo98
WRZ from 1:4000 map in Fig. 2 of Ch98 (only part of
SRL). Surface ruptures coincide with or are very close
to preexisting MF scarps.
Not quantified
1995 Egion,
Greece; 6.2
Egion N 7.2 0.03 (v)1–60 KD96 En echelon ruptures at the western termination of MF.
Separation between segments is from 80 to 150 m.
100 Not quantified 0
2006 Machaze,
Mozambique;
7.0
Machaze NL
(1)
>15
(30–40)
2.05 (v),
0.7 (h)
1–140 (2) FB06 WRZ from ∼1:10,000 maps in Fig. 3 of FB06 (only part
of SRL). (1) The fault is drawn with a dextral
component on the map, but is indicated as left in the
text and photo. (2) The maximum value in the northern
strands.
100 Not quantified
2009 L’Aquila,
Italy; 6.3
Paganica,
San Gregorio
N 13 0.12 (v)1–140 This
paper
The most constrained data along the Paganica fault. 100 98.7
(Paganica)
0
*Kin., kinematics; N, normal faulting (dip-slip); NR, normal (prevailing) right-oblique; RN, right-lateral (prevailing) normal-oblique; NL, normal (prevailing) left-oblique; LN, left-lateral (prevailing) normal-
oblique; (1) refers to notes for the 2006 Machaze earthquake.
†MD, maximum surface displacement; (v), vertical; (h), horizontal.
‡1, single scarp along the reactivated main fault; (1) and (2) refer to notes.
§Ref., references; GG99, Galadini and Galli (1999); Wa84, Wallace (1984); Be91, Bellier et al. (1991); Gi57, Gianella (1957); Ca96, Caskey et al. (1996); Wi62, Witkind et al. (1962); Ta71, Tasdemiroglu (1971);
Cl76, Clark et al. (1976); PV90, Pantosti and Valensise (1990); Pa93, Panayotis et al. (1983); PC04, Pavlides and Caputo (2004); Ja82, Jackson et al. (1982); Cr87, Crone et al. (1987); Ly88, Lyon-Caen et al. (1988);
Be89, Beanland et al. (1989); Ch98, Chatzipetros et al. (1998); Mo98, Mountrakis et al. (1998); KD96, Koukouvelas and Doutsos (1996); FB06, Fenton and Bommer (2006).
∥SR, surface ruptures; MF, main fault; HW, hanging wall; FW, footwall.
#% SRL HW ≤150, percentage of SRLwith surface ruptures in the hanging wall at distances ≤150 m from the trace of the main fault (MF); % SRL HW ≤40, percentage of SRL with surface ruptures in the hanging
wall at distances ≤40 m from the trace of MF; % SRL FW, percentage of SRL with surface ruptures along footwall splays. The question mark (?) indicates uncertainty.
**(1) and (2) for the 1981 Corinth earthquakes refer to different faults reactivated during the seismic sequence.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 929
Figure 9. (a, b) Frequency distribution histograms of the WRZ measured for the earthquakes reported in Table 1. The positive and
negative WRZ values refer to data measured on the hanging wall and footwall of the main fault (MF), respectively. (c–f) Examples of large
WRZ data measured at major geometric complexities of the main fault for dip-slip (c, f) and oblique-slip (d, e) ruptures (redrawn from Caskey
et al., 1996 [c, d, e] and Crone et al., 1987 [f]).
930 P. Boncio, P. Galli, G. Naso, and A. Pizzi
complexities and produces wider rupture zones. However,
considering the reactivated fault segments individually, a po-
sitive relation between WRZ and magnitude is not evident,
and the rupture zone is systematically narrow, mostly within
∼150 m from the main fault trace (Fig. 9b).
Delineating EFZs and Ss for Apennine-Type
Normal Faults
The model in Figure 11a–dmight help in defining
general criteria for delineating EFZs and Ss for active Apen-
nine-type dip-slip normal faults. The model is inspired by the
L’Aquila 2009 earthquake, but it also benefits from a com-
parison with WRZ data from normal-faulting earthquakes
worldwide.
We suggest that both the EFZ and the Sshould be asym-
metrically shaped around the trace of the active fault, with a
wider zone on the hanging wall than on the footwall. This
accounts for the observation that the fault ruptures mostly
occur along the fault trace and in the hanging wall. This
is also in agreement with the geologic observation that the
deformation associated with an inclined dip-slip fault mostly
concentrates in the hanging wall block (e.g., Fig. 9a,b).
The width of the main deformation zone observed along
the Paganica fault can be used for defining the setbacks. The
proposed setback on the hanging wall (Shw in Fig. 11a)is
40 m, which includes the belt of coseismic faulting and frac-
turing (30–35 m) plus an estimated mapping error of 5 m.
This value is in agreement with WRZ data collected world-
wide (Fig. 9b). In the footwall, we propose a narrower set-
back (Sfw in Fig. 11a). The proposed minimum width of the
Figure 10. Plot of WRZ versus magnitude for earthquakes re-
ported in Table 1. Note that large WRZ values (>150 m) mostly
come from broad deformation zones at major geometric complex-
ities of the main fault.
Figure 11. Proposal of general criteria, based on the Paganica and San Gregorio normal faults, for shaping and sizing the EFZ and Son
both the hanging wall (hw) and footwall (fw) of active normal faults. (a) Known fault trace. (b) Fault trace bracketed within a zone of
geological uncertainty (g.u.). (c) Uncertain fault trace. (d) Map view.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 931
setback in the footwall is 15 m to account for uncertainties in
the fault location, such as faults that appear as degraded
scarps in poorly consolidated deposits. The value of 15 m
is in line with minimum setbacks adopted in the United
States (Bryant and Hart, 2007; see also Batatian, 2002,
Recommendations for Fault Setbacks).
The EFZ (Fig. 11a) should be wider than the setback.
The EFZ should be a zone mapped prior to an earthquake
during standard seismic microzonation work, and it should
include all the reasonably inferred fault-rupture hazards, both
on the main fault and the possible active branches of the main
fault. Detailed investigation prior to building will define any
additional hazard due to active branches. If active branches
are present, they will be traced in detail, and the appropriate
setbacks will be determined. Using the history of the Paga-
nica fault, the associated EFZ must include both the main
deformation zone and the reactivated synthetic splays located
farther from the fault trace (120–140 m). Therefore, the edge
of the EFZ on the hanging wall should be located at least
150 m from the trace of the main fault (EFZhw in Fig. 11a,
d). Statistical analysis of WRZ data for normal-faulting earth-
quakes, particularly the dip-slip earthquakes, supports 150 m
as a reasonable width for the zone within which structures
associated with the main fault might be activated. An EFZhw
of 150 m does not assure that ruptures cannot occur outside
the zone. Nevertheless, the data collected here suggest that
ruptures occurring far from the main fault reactivate preex-
isting faults that may be recognized by field mapping. If de-
tailed field analysis, necessary for this kind of evaluation,
identify associated structures located far from the main fault,
they can be treated as individual faults, each with its own EFZ
and setback. Particular attention must be paid to geometric or
structural complexities between adjacent fault segments. If
historic or paleoseismic data suggests that the rupture can
cross the zone of fault complexity during a single large earth-
quake, a wider EFZ aimed at covering the entire zone of com-
plexity (e.g., the overlapping zone in a stepover) might be
adopted (Fig. 11d).
On the footwall, it is difficult to establish the appropriate
width of the EFZ. Footwall splays are often absent or spora-
dic features and may occur over a large range of distances
from the main fault (e.g., Table 1and Fig. 9). Detailed field
analysis might recognize active footwall splays that can be
treated as individual faults with their own EFZs and setbacks.
For a mapped active fault, we propose an EFZ width for the
footwall (EFZfw in Fig. 11a) that is aimed at incorporating
possible uncertainties in the width of the fault zone, rather
than at incorporating the probability of occurrence of foot-
wall splays. With this goal, a width of 30 m is probably suf-
ficient for a well mapped fault (e.g., 1:5000 scale). Although
this choice is rather arbitrary and based on our field mapping
experience rather than on rigorous constraints, it seems rea-
sonable, provided that the mapping is of sufficient detail.
In cases of uncertain fault traces, using a zone of geo-
logic uncertainty could be useful. Figure 11b shows an ex-
ample, based on Apennine-type normal faults, of a fault trace
bracketed within a zone of geologic uncertainty defined by
all geologic and geomorphic data acquired during detailed
field mapping (e.g., during seismic microzonation studies;
g.u. in Fig. 11b). The resulting EFZ is the sum of the hanging
wall and footwall EFZs, described previously, and the width
of the zone of geologic uncertainty. In cases where the width
of the area of geologic uncertainty cannot be assessed (e.g.,
areas with flat topography; Fig. 11c), one could adopt a sym-
metric EFZ centered on the most likely fault trace and having
the maximum width on both the hanging wall and footwall
(e.g., 150 150 m).
Finally, considering both the data gathered during the
2009 L’Aquila earthquake and their comparison with WRZ
data collected for several normal-faulting earthquakes, we
suggest that the shape and size of the EFZs and Ss proposed
here are sufficiently conservative for normal faults in an
Italian-type seismotectonic context. In Figure 12, there are
examples of applications to the Paganica fault, studied here,
and to the active Sulmona normal fault, one of the most
hazardous faults in central Italy (Galadini and Galli, 2000;
Pace et al., 2006).
Conclusions
Mapping the coseismic surface faulting associated with
the 2009 L’Aquila normal-faulting earthquake in central Italy
provides insights on the general criteria for shaping zones of
SFRH along active normal faults. This is useful for (1) addres-
sing policies aimed at mitigating the SFRH in Italy and in
areas with a comparable seismotectonic setting that do not
already have specific regulations, and (2) providing a case
study that might help in defining basic criteria for determin-
ing EFZs and Ss along active normal faults, even for countries
or local authorities that already have regulations for SFRH.
Both EFZ and Sare needed to adequately account for the
likely SFRH across an active normal fault. The EFZ should
include the main fault and the possible active branches of the
main fault. Following the A–P(Bryan and Hart, 2007), the
EFZ should be a regulatory zone within which detailed inves-
tigations are required prior to building structures for human
occupancy. The Sis the fault-avoidance zone within which
critical facilities and structures for human occupancy cannot
be built. While detailed geological investigations prior to
building structures for human occupancy can better define
the local hazard, minimum requirements for the shape and
width of the EFZ and Sare a necessary starting point.
The 2009 L’Aquila earthquake is one of the smaller
earthquakes (Mw6.3) with ground surface rupture, offering
the opportunity to set these minimum requirements for nor-
mal faults on the basis of direct observations. The conclu-
sions based on these observations are strengthened by the
comparison with surface rupture data collected for several
normal-faulting earthquakes worldwide.
Both the EFZ and Sshould be asymmetrically shaped
around the trace of the active fault. An Sof 15 m is a reason-
able value for the footwall, provided that the fault is traced on
932 P. Boncio, P. Galli, G. Naso, and A. Pizzi
a detailed map (e.g., 1:5000 scale). On the hanging wall, a
more precautionary Sof 40 m is suggested. Observation sug-
gests a minimum width of 150 m for the EFZ on the hanging
wall, in order to account for possible active branches of the
main fault. A minimum EFZ width of 30 m on the footwall
seems to us sufficient for faults mapped in detail.
Rupture zone widths observed globally during moder-
ate-to-large normal-faulting earthquakes indicate that the
choice of an asymmetric zone around the main fault is rea-
sonable, and the proposed widths of the Sand EFZ are sta-
tistically reasonable. These data also suggest that other
factors must be considered, such as fault kinematics and fault
complexities. The WRZ seems to be systematically wider for
oblique-slip faults, indicating that the proposed values are
particularly suited for dip-slip normal faults. During large
earthquakes, the rupture can cross large (i.e., kilometer-scale)
geometric complexities, producing rupture zones much
wider than for simple fault traces. This suggests that the
width of the EFZ must be flexible, widening in zones of fault
complexity (e.g., as large as the separation zone between en
echelon segments).
The criteria and examples presented here may be useful
in evaluating the potential for SFRH. On the other hand, they
cannot be considered operational instruments (e.g., in official
construction regulations) until they are at least discussed
with other specialists involved in the problem such as engi-
neers and planners. Moreover, we believe that our geologic
criteria for shaping SFRH zones could provide a starting point
for a risk-based approach (i.e., an approach that takes into
account the type of structure/building that may be impacted
by rupture and the fault recurrence interval), similar to the
one adopted in New Zealand (Kerr et al., 2003).
We hope this scientific contribution will promote discus-
sion, hopefully with operational implications.
Data and Resources
Surface rupture data for the 2009 L’Aquila earthquake
are from original field work performed by us in the epicentral
area. Detailed descriptions, precise geographic locations of
single observations, and additional tectonic information/in-
terpretations are available in published papers listed in Re-
ferences. The elaboration of surface rupture data in terms of
earthquake fault zoning is original work that evolved during
seismic microzonation-oriented activities performed by us
after the earthquake. Global surface rupture data of nor-
mal-faulting earthquakes are from a compilation and synth-
esis of published papers reported in Table 1and listed in the
Figure 12. Application of the EFZ and Sto (a) the Paganica fault, activated during the 2009 L’Aquila earthquake, and (b) the active
Sulmona normal fault near Roccacasale (location indicated in the inset; field view of the fault plane at the bottom right). The color version of
this figure is available only in the electronic edition.
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 933
References section. Geologic data were managed mostly
with the software ArcGIS 9.3 by ESRI (licensed to the
G. D’Annunzio University of Chieti-Pescara).
Acknowledgments
This work was funded by the Ministero dell’Istruzione, dell’Università
e della Ricerca (MIUR; grants to P. Boncio and A. Pizzi) and by the Italian
Department of Civil Protection (funds for seismic microzonation of L’Aquila
after the 6 April 2009 earthquake; see www.protezionecivile.it). We ac-
knowledge Associate Editor B. Sherrod and three anonymous referees of
BSSA for the constructive revision of the manuscript.
References
Anzidei, M., E. Boschi, V. Cannelli, R. Devoti, A. Esposito, A. Galvani, D.
Melini, G. Pietrantonio, F. Riguzzi, V. Sepe, and E. Serpelloni (2009).
Coseismic deformation of the destructive April 6, 2009 L’Aquila earth-
quake (central Italy) from GPS data, Geophys. Res. Lett. 36, L17307,
doi 10.1029/2009GL039145.
APAT (2005). Carta Geologica d’Italia, F. 359 L’Aquila, scale 1:50,000,
APAT (Servizio Geologico d’Italia) and Regione Abruzzo, S.EL.CA,
Firenze.
Atzori, S., I. Hunstad, M. Chini, S. Salvi, C. Tolomei, C. Bignami, S. Stra-
mondo, E. Trasatti, A. Antonioli, and E. Boschi (2009). Finite
fault inversion of DInSAR coseismic displacement of the 2009 L’A-
quila earthquake (central Italy), Geophys. Res. Lett. 36, L15305 doi
10.1029/2009GL039293.
Bagnaia, R., A. D’Epifanio, and S. Sylos Labini (1992). Aquila and
Subequan basins: An example of Quaternary evolution in central
Apennines, Italy, Quaternaria Nova 2, 187–209.
Batatian, D.(2002). Minimum standards for surface fault rupture
hazard studies, in Salt Lake County Geologic Hazards Ordinance,
available at www.pwpds.slco.org/zoning/pdf/geologichazards/
AppAfaultReportMinStds.pdf.
Beanland, S., K. R. Berryman, and G. H. Blick (1989). Geological inves-
tigations of the 1987 Edgecumbe earthquake, New Zealand, New Zeal.
J. Geol. Geophys. 32, 73–91.
Bellier, O., J. F. Dumont, M. Sébrier, and J. L. Mercier (1991). Geological
constraints on the kinematics and fault-plane solution of the Quiches
fault zone reactivated during the 10 November 1946 Ancash earth-
quake, northern Peru, Bull. Seismol. Soc. Am. 81, 468–490.
Bertini, T., and C. Bosi (1993). La tettonica quaternaria della conca di Fossa
(L’Aquila), Il Quaternario 6, 293–314.
Blumetti, A. M., F. Dramis, and A. M. Michetti (1993). Fault-generated
mountain fronts in the central Apennines (Central Italy): Geomorpho-
logical features and seismotectonic implications, Earth Surf. Proc.
Land. 18, 203–223, doi 10.1002/esp.3290180304.
Boncio, P., F. Brozzetti, and G. Lavecchia (2000). Architecture and seismo-
tectonics of a regional low-angle normal fault zone in central Italy,
Tectonics 19, 1038–1055, doi 10.1029/2000TC900023.
Boncio, P., G. Lavecchia, G. Milana, and B. Rozzi (2004). Seismogenesis in
Central Apennines, Italy: An integrated analysis of minor earthquake
sequences and structural data in the Amatrice-Campotosto area, Ann.
Geophysics 47, 1723–1742.
Boncio, P., G. Lavecchia, and B. Pace (2004). Defining a model of 3D seis-
mogenic sources for seismic hazard assessment applications: The case
of central Apennines (Italy), J. Seismol. 8, 407–425, doi 10.1023/B:
JOSE.0000038449.78801.05.
Boncio, P., A. Pizzi, F. Brozzetti, G. Pomposo, G. Lavecchia, D. Di Naccio,
and F. Ferrarini (2010). Coseismic ground deformation of the 6 April
2009 L’Aquila earthquake (central Italy, Mw6.3), Geophys. Res. Lett.
37, L06308, doi 10.1029/2010GL042807.
Boncio, P., A. Pizzi, G. Cavuoto, M. Mancini, T. Piacentini, E. Miccadei,
G. P. Cavinato, S. Piscitelli, A. Giocoli, G. Ferretti, R. De Ferrari, M.
R. Gallipoli, M. Mucciarelli, V. Di Fiore, A. Franceschini, F. Pergalani,
and G. Naso, and Working Group Macroarea 3 (2011). Geological and
geophysical characterisation of the Paganica—San Gregorio area after
the April 6, 2009 L’Aquila earthquake (Mw6.3, central Italy): Impli-
cations for site response, Boll. Geofis. Teor. Appl. 52, 491–512 doi
10.4430/bgta0014.
Bosi, C., and T. Bertini (1970). Geologia della media valle dell’Aterno,
Mem. Soc. Geol. Ital. 9, 719–777.
Bryant, W. A., and E. W. Hart (2007). Fault-rupture hazard zones in Cali-
fornia: Alquist–Priolo earthquake fault zoning act with index to earth-
quake fault zones maps. Calif. Geol. Surv., Spec. Pub. 42, 41 pp.
Calamita, F., and A. Pizzi (1994). Recent and active extensional tectonics in
the southern Umbro–Marchean Apennines (Central Italy), Mem. Soc.
Geol. Ital. 48, 541–548.
Caskey, S. J., S. G. Wesnousky, P. Zhang, and D. B. Slemmons (1996). Sur-
face faulting of the 1954 Fairview Peak (Ms7.2) and Dixie Valley
(Ms6.8) earthquakes, central Nevada, Bull. Seismol. Soc. Am. 86,
761–787.
Chatzipetros, A. A., S. B. Pavlides, and D. M. Mountrakis (1998). Under-
standing the 13 May 1995 western Macedonia earthquake: A paleo-
seismological approach, J. Geodynamics 26, 327–339, doi 10.1016/
S0264-3707(97)00069-0.
Chiarabba, C., A. Amato, M. Anselmi, P. Baccheschi, I. Bianchi, M.
Cattaneo, G. Cecere, L. Chiaraluce, M. G. Ciaccio, P. De Gori, G.
De Luca, M. Di Bona, R. Di Stefano, L. Faenza, A. Govoni, L.
Improta, F. P. Lucente, A. Marchetti, L. Margheriti, F. Mele, A. Mi-
chelini, G. Monachesi, M. Moretti, M. Pastori, N. Piana Agostinetti, D.
Piccinini, P. Roselli, D. Seccia, and L. Valoroso (2009). The 2009 L’A-
quila (central Italy) Mw6.3 earthquake: Main shock and aftershocks,
Geophys. Res. Lett. 36, L18308, doi 10.1029/2009GL039627.
Cinti, F. R., D. Pantosti, P. M. De Martini, S. Pucci, R. Civico, S. Pierdo-
minici, L. Cucci, C. A. Brunori, S. Pinzi, and A. Patera (2011). Evi-
dence for surface faulting events along the Paganica fault prior to the 6
April 2009 L’Aquila earthquake (central Italy), J. Geophys. Res. 116,
21, doi 10.1029/2010JB007988.
Cirella, A., A. Piatanesi, M. Cocco, E. Tinti, L. Scognamiglio, A. Michelini,
A. Lomax, and E. Boschi (2009). Rupture history of the 2009 L’Aquila
(Italy) earthquake from non-linear joint inversion of strong motion
and GPS data, Geophys. Res. Lett. 36, L19304 doi 10.1029/
2009GL039795.
Clark, M. M., R. V. Sharp, R. O. Castle, and P. W. Harsh (1976). Surface
faulting near Lake Oroville, California in August, 1975, Bull. Seismol.
Soc. Am. 66, 1101–1110.
Crone, A. J., M. N. Machette, M. G. Bonilla, J. J. Lienkaemper, K. L. Pierce,
W. E. Scott, and R. C. Bucknam (1987). Surface faulting accompany-
ing the Borah Peak earthquake and segmentation of the Lost River
fault, central Idaho, Bull. Seismol. Soc. Am. 77, 739–770.
DISS Working Group (2010). Database of Individual Seismogenic Sources
(DISS), Version 3.1.1: A compilation of potential sources for earth-
quakes larger than M5.5 in Italy and surrounding areas, http://diss
.rm.ingv.it/diss/,© INGV 2010—Istituto Nazionale di Geofisica e Vul-
canologia (last accessed November 2011).
Eurocode 8 (2003). Design of structures for earthquake resistance: Founda-
tions, retaining structures and geotechnical aspects, European Com-
mittee for Standardization (CEN), Brussels, http://www.cen.eu/
cenorm/homepage.htm (last accessed November 2011).
EMERGEO Working Group (2009). Evidence for surface rupture
associated with the Mw6.3 L’Aquila earthquake sequence of April
2009 (central Italy), Terra Nova 22, 43–51, doi 10.1111/j.1365-
3121.2009.00915.x.
Falcucci, E., S. Gori, E. Peronace, G. Fubelli, M. Moro, M. Saroli, B. Giac-
cio, P. Messina, G. Naso, G. Scardia, A. Sposato, M. Voltaggio, P.
Galli, and F. Galadini (2009). The Paganica fault and surface coseismic
ruptures caused by the 6 April 2009 earthquake (L’Aquila, Central
Italy), Seismol. Res. Lett. 80, 940–950, doi 10.1785/gssrl.80.6.940.
Fenton, C. H., and J. J. Bommer (2006). The Mw7 Machaze, Mozambique,
earthquake of 23 February 2006, Seismol. Res. Lett. 77, 426–439, doi
10.1785/gssrl.77.4.426.
934 P. Boncio, P. Galli, G. Naso, and A. Pizzi
Galadini, F., and P. Galli (1999). The Holocene paleoearthquakes on the
1915 Avezzano earthquake faults (central Italy): Implications for active
tectonics in Central Apennines, Tectonophysics 308, 143–170, doi
10.1016/S0040-1951(99)00091-8.
Galadini, F., and P. Galli (2000). Active tectonics in the Central Apennines
(Italy)—Input data for seismic hazard assessment, Natural Hazards
22, 225–268, doi 10.1023/A:1008149531980.
Galli, P., R. Camassi, R. Azzaro, F. Bernardini, S. Castenetto, D. Molin, E.
Peronace, A. Rossi, M. Vecchi, and A. Tertulliani (2009). II terremoto
aquilano del 6 aprile 2009: Rilievo macrosismico, effetti di superificie
ed implicazioni sismotettoniche, Il Quaternario 22, 235–246.
Galli, P., F. Galadini, and F. Calzoni (2005). Surface faulting in Norcia
(central Italy): A “paleoseismological perspective”,Tectonophysics
403, 117–130, doi 10.1016/j.tecto.2005.04.003.
Galli, P., F. Galadini, and D. Pantosti (2008). Twenty years of paleoseismol-
ogy in Italy, Earth Sci. Rev. 88, 89–117 doi 10.1016/j.earscir-
ev.2008.01.001.
Galli, P., B. Giaccio, and P. Messina (2010). The 2009 central Italy earth-
quake seen through 0.5 Myr-long tectonic history of the L’Aquila
faults system, Quat. Sci. Rev. 29, 3768–3789, doi 10.1016/j.quascir-
ev.2010.08.018.
Gianella, V. P. (1957). Earthquake and faulting, Fort Sage mountains,
California, December 1950, Bull. Seismol. Soc. Am. 47, 173–177.
Jackson, J. A., J. Gagnepain, G. Houseman, G. C. P. King, P. Papadimitriou,
C. Soufleris, and J. Virieux (1982). Seismicity, normal faulting, and the
geomorphological development of the Gulf of Corinth (Greece): The
Corinth earthquakes of February and March 1981, Earth Plan. Sci.
Lett. 57, 377–397, doi 10.1016/0012-821X(82)90158-3.
Kaven, J. O., and S. J. Martel (2007). Growth of surface-breaching normal
faults as a three-dimensional fracturing process, J. Struct. Geol. 29,
1463–1476, doi 10.1016/j.jsg.2007.05.007.
Kerr, J., S. Nathan,R. Van Dissen, P. Webb,D. Brunsdon, and A. King (2003).
Planning for the development of land on or close to active faults, A
guideline to assist resource management planners in New Zealand,
http://www.mfe.govt.nz/publications/rma/planning‑development
‑active‑faults‑dec04/index.html, Ministry for the Environment, Well-
ington, New Zealand (last accessed June 2011).
Koukouvelas, I. K., and T. T. Doutsos (1996). Implications of structural
segmentation during earthquakes: The 1995 Egion earthquake, Gulf
of Corinth, Greece, J. Struct. Geol. 18, 1381–1388, doi 10.1016/
S0191-8141(96)00071-5.
Lavecchia, G., F. Brozzetti, M. R. Barchi, J. Keller, and M. Menichetti
(1994). Seismotectonic zoning in east-central Italy deduced from
the analysis of the Neogene to present deformations and related stress
fields, Soc. Geol. Am. Bull. 106, 1107–1120.
Lyon-Caen, H., R. Armijo, J. Drakopoulos, J. Baskoutass, N. Delibassis, R.
Gaulon, V. Kouskouna, J. Latoussakis, K. Makropoulos, P. Papadimi-
triou, D. Papanastassiou, and J. Pedotti (1988). The 1986 Kalamata
(South Peloponnesus) earthquake: Detailed study of a normal fault,
evidences for east-west extension in the Hellenic Arc, J. Geophys.
Res. 93, 14,967–15,000, doi 10.1029/JB093iB12p14967.
Martel, S. J., and J. Langley (2006). Propagation of normal faults to the sur-
face in basalt, Koae fault system, Hawaii, J. Struct. Geol. 28, 2123–
2143, doi 10.1016/j.jsg.2005.12.004.
Mountrakis, D., S. Pavlides, N. Zouros, Th. Astaras, and A. Chatzipetros
(1998). Seismic fault geometry and kinematics of the 13 May 1995
western Macedonia (Greece) earthquake, J. Geodyn. 26, 175–196,
doi 10.1016/S0264-3707(97)00082-3.
NTC (2008). Norme Tecniche per le Costruzioni. DM 14 gennaio 2008,
Gazzetta Ufficiale, n. 29 del 4 febbraio 2008, Supplemento Ordinario
n. 30, www.cslp.it, Istituto Poligrafico e Zecca dello Stato, Roma (last
accessed November 2011).
Pace, B., L. Peruzza, G. Lavecchia, and P. Boncio (2006). Layered seismo-
genic source model and probabilistic seismic-hazard analyses in
central Italy, Bull. Seismol. Soc. Am. 96, 107–132, doi 10.1785/
0120040231.
Panayotis, G. C., R. T. Norman, G. E. Brandow, and J. O. Jirsa (1983). The
Central Greece Earthquakes of February–March 1981: A Reconnais-
sance and Engineering Report,National Academy Press, Washington,
1–25.
Pantosti, D., and L. Valensise (1990). Faulting mechanism and complexity of
the November 23, 1980 Campania-Basilicata earthquake, inferred
from surface observations, J. Geophys. Res. 95, 15,319–15,342.
Pavlides, S., and R. Caputo (2004). Magnitude versus faults’surface para-
meters: Quantitative relationships from the Aegean region, Tectono-
physics 380, 159–188, doi 10.1016/j.tecto.2003.09.019.
Pizzi, A., F. Calamita, M. Coltorti, and P. Pieruccini (2002). Quaternary
normal faults, intramontane basins and seismicity in the Umbria-
Marche-Abruzzi Apennine ridge (Italy): Contribution of neotectonic
analysis to seismic hazard assessment, Boll. Soc. Geol. It., Spec. Publ.
1, 923–929.
Pondrelli, S., S. Salimbeni, A. Morelli, G. Ekstrom, M. Olivieri, and E.
Boschi (2010). Seismic moment tensors of the April 2009, L’Aquila
(Central Italy), earthquake sequence, Geophys. J. Int. 180,, 238–22,
doi 10.1111/j.1365-246X.2009.04418.x.
Tasdemiroglu, M. (1971). The 1970 Gediz earthquake in western Anatolia,
Turkey, Bull. Seismol. Soc. Am. 61, 1507–1527.
Vezzani, L., and F. Ghisetti (1998). Carta geologica dell’Abruzzo, S.EL.CA,
Firenze, scale1:100,000.
Vittori, E., P. Di Manna, A. M. Blumetti, V. Comerci, L. Guerrieri, E. Espo-
sito, A. M. Michetti, S. Porfido, L. Piccardi, G. P. Roberts, A. Berlus-
coni, F. Livio, G. Sileo, M. Wilkinson, K. J. W. McCaffrey, R. J.
Phillips, and P. A. Cowie (2011). Surface faulting of the 6 April
2009 Mw6.3 L’Aquila earthquake in central Italy, Bull. Seismol.
Soc. Am. 101, 1507–1530, doi 10.1785/0120100140.
Wallace, R. E. (1984). Fault scarps formed during the earthquake of October
2, 1915, Pleasant Valley, Nevada and some tectonic implications, U.S.
Geol. Surv. Profess. Pap. 1274-A,1–33.
Walters, R. J., J. R. Elliott, N. D’Agostino, P. C. England, I. Hunstad, J. A.
Jackson, B. Parsons, R. J. Phillips, and G. Roberts (2009). The 2009
L’Aquila earthquake (central Italy): A source mechanism and implica-
tions for seismic hazard, Geophys. Res. Lett. 36, L17312, doi 10.1029/
2009GL039337.
Witkind, I. J., W. B. Myers, J. B. Hadley, W. Hamilton, and G. D. Fraser
(1962). Geologic features of the earthquake at Hebgen Lake, Montana,
August 17, 1959, Bull. Seismol. Soc. Am. 52, 163–180.
Working Group CPTI (2004). Catalogo parametrico dei Terremoti Italiani,
versione 2004 (CPTI04), INGV-Bologna: available on-line at http://
emidius.mi.ingv.it/CPTI.
Working Group MS (2008). Indirizzi e Criteri per la Microzonazione
Sismica. Conferenza delle Regioni e delle Province autonome,
Dipartimento della protezione civile, Roma, 3 vol. and Cd-rom.
Università di Chieti-Pescara
Via dei Vestini, 30-66013
Chieti scalo, Italy
pboncio@unich.it
pizzi@unich.it
(P.B., A.P.)
Dipartimento della Protezione
Civile-Ufficio Rischio Sismico
Via Vitorchiano, 4-00189
Roma, Italy
Paolo.Galli@protezionecivile.it
Giuseppe.Naso@protezionecivile.it
(P.G., G.N.)
Manuscript received 2 November 2010
Zoning Surface Rupture Hazard along Normal Faults: The L’Aquila and Other Earthquakes 935
