Co-seismic displacements associated to the Molise (Southern Italy) earthquake sequence of October–November 2002 inferred from GPS measurements
Introduction
On October 31, 2002 at 10:32 GMT an earthquake (Ml = 5.4, Mw = 5.7, M0 = 4.3 ⁎ 1017 Nm) struck eastern Molise in Southern Italy, followed on the next day, the November 1, 2002 at 15:08 GMT, by another main shock of comparable energy (Ml = 5.0, Mw = 5.7, M0 = 4.6 ⁎ 1017 Nm) localized a few kilometers westwards (Fig. 1). The hypocentral depths are 18–22 km depth for both main shocks (Chiarabba et al., 2005).
The centroid moment tensor focal solutions of the two main shocks (http://mednet.ingv.it/events/Welcome.html, http://www.seismology.harvard.edu/projects/CMT) identify two vertical planes with an almost pure strike-slip kinematics. The aftershock sequence (Fig. 1), recorded by local seismic network deployed by INGV (Istituto Nazionale di Geofisica Vulcanologia) in the days following the main shocks, defines an E–W trending focal volume, identifying the E–W plane as the most likely fault rupture, clearly suggests that the focal solution for both main shocks is the E–W trending vertical plane with right-lateral strike-slip kinematics. The aftershocks vertical distribution (Fig. 1), between 20 and 5 km depth, is shallower than the deep location of both main shocks (Chiarabba et al., 2005).
The kinematics and the deep location of this earthquake are unusual in the framework of the active tectonic setting of the southern Apennines, apart from the 1990 Potenza seismic sequence. This earthquake is similar to the Molise sequence as far as its focal depth (between 15 and 25 km), its intermediate magnitudes (Ml = 5.2) and its focal mechanism kinematics (Ekstrom, 1994). Moreover in this area no local seismogenetic structures have been recognized according to seismic catalogue and on the basis of geological and surface faulting studies (Molin, 1987, Galadini et al., 2000, Galli and Molin, 2004). Therefore the characterization of the seismic source using the contribution of independent data, such as geodetic measurements of superficial co-seismic displacements, is particularly significant for this earthquake in order to understand its meaning in terms of the seismogenic framework of this part of Apennines.
A GPS survey, conducted by Department of Civil Protection–National Seismic Survey (DPC–SSN) and the Italian Agency for Environment Protection and Technical Services (APAT), in May 2001 on a geodetic network in Campania and Molise regions, makes up part of the data set used in this study. Other GPS measurements of geodetic benchmarks in the epicentral area performed before the earthquake by the INGV, complete the pre-earthquake data set. The GPS re-measurements on geodetic benchmarks located in the epicentral area (within 25 km from the seismic epicenter) have been performed jointly by DPC–SSN, APAT and INGV a few days after two main shocks in order to infer the co-seismic displacements associated to the earthquake.
In this paper we present the co-seismic displacement vectors associated to the 2002 Molise seismic sequence inferred from the results of processing of GPS measurements and an elastic dislocation modeling (Okada, 1985) of the fault rupture associated to the two main shocks. Sensitivity tests have been performed in order to infer the slip distribution at depth and to define which focal depth, assumed in the seismic source modelling, better fits observed displacements. The seismotectonic hypothesis of present paper is to link the Molise seismic sequence to the seismicity that represents the expression of the decoupling of two blocks within the Adriatic microplate corresponding to the Gargano–Dubrovnik deformation zone (Westaway, 1990, Calais et al., 2002, D'Agostino et al., 2006).
Section snippets
Seismotectonic framework
The epicentral area of Molise 2002 earthquake sequence is located between the axis of Apennines chain westwards and the Apulian foreland towards East; this area is placed immediately inside the Apennine buried frontal ramp whose compressive activity is sealed by marine uppermost Pliocene–Lower Pleistocene sequences, gently tilted towards ENE (Cinque et al., 1993, Patacca and Scandone, 2004).
The axial southern Apennines belt is characterized by an intense seismicity (CPTI, 1999) with high energy
The earthquakes of October 31, 2002
The aftershock sequence (Fig. 1), recorded by local seismic network deployed by INGV (Istituto Nazionale di Geofisica Vulcanologia) in the days following the main shocks, defines an E–W trending focal volume, identifying the E–W plane as the most likely fault rupture with the majority of events ranging from 5 to 20 km of depth. Both main shocks are placed in the lower portion of the focal volume defined by the location of aftershocks sequence suggesting that rupture directivity had a
Pre-earthquake GPS data
The pre-earthquake GPS data derives from two different sources: a campaign performed in 2001 by DPC–SSN and APAT (at that moment National Geologic Service) and other GPS observations carried out by INGV in 2001 and in July 2002.
The 2001 campaign consisted of GPS observations on vertexes of a sub-network of the IGM95 (Surace, 1997) that is the first order national geodetic GPS network firstly measured in 1994 by Istituto Geografico Militare (IGM). This geodetic network covers the whole Italian
Data analysis
All the GPS data analyzed to estimate the site coordinates before and after the earthquake have been checked using the package distributed by UNAVCO (www.unavco.ucar.edu/software/teqc) in order to perform a quality check of GPS observation files (rinex files). Observation data set satisfied quality check.
Since the expected surface deformation induced by main shocks of the Molise seismic sequence are of the order of the GPS reliability, we choose to process collected data by use of two codes
Seismic source model
Following the recognition of geodetic displacements associated to the earthquake, we used GPS data results to contribute to the seismic source characterization with a methodology comparable to the one used in Hunstad et al. (1999) for 1997 Umbria Marche earthquake. Therefore we modelled the earthquake slip on the basis of available seismological data as constrains to the model assuming a finite source dislocation within an elastic half-space (Okada, 1985). A complete inversion of the data for
Conclusion
Using two different processing softwares and strategies to analyze GPS data we observed a comparable superficial displacement field statistically significant for a time span that comprises both main shocks of Molise 2002 earthquake. Even for a deep earthquake with a moderate magnitude, the GPS re-measurement in the epicentral area allowed to infer a co-seismic strain field requiring a geodetic moment release similar to the seismic one inferred from seismological data. The axes of horizontal
Acknowledgements
We thank Riccardo Bianconi, Giuseppe Casula, Alessandra Esposito, Valeria Eulilli, Piera Gambino, Fabiana Loddo, Amedeo Lucidi, Domenico Matarazzo, Franco Palma, Claudio Pulsinelli, Franco Valli, Francesco Vullo for their help during the field work with GPS. We also thank Martin Vallée for useful discussions on the source model and Paolo Galli for discussions on the seismotectonic framework and for providing Fig. 2, Fig. 3.
References (45)
- et al.
The 2002 Molise earthquake sequence: what can we learn about the tectonics of southern Italy?
Tectonophysics
(2005) Accuracy of modern global earthquake catalogs
Phys. Earth Planet. Inter.
(2003)- et al.
Three-dimensional strain tensor estimation by GPS observations: methodological aspects and geophysical applications
J. Geodyn.
(2004) Present-day kinematics of the plate boundary zone between Africa and Europe, from the Azores to the Aegean
Earth Planet. Sci. Lett.
(1990)- et al.
Active tectonics of the Adriatic region
Geophys. J. R. Astron. Soc.
(1987) - et al.
Insights into present-day crustal motion in the central Mediterranean area from GPS surveys
Geophys. J. Int.
(2001) - et al.
Bernese Software 4.2
(2000) - et al.
The large Salcito Landslide triggered by the 2002 Molise, Italy, Earthquake
Earthq. Spectra
(2004) - et al.
Applying thick-skinned tectonic models to the Apennine thrust belt of Italy: limitations and implications
- et al.
Current strain regime in the Western Alps from continuous Global Positioning System measurements, 1996–2001
Geology
(2002)
Mainshocks and aftershocks of the 2002 Molise seismic sequence, southern Italy
J. Seismol.
Wrench zone reactivation in the Adriatic Block: the example of the Mattinata fault system (SE Italy)
Boll. Soc. Geol. Ital.
Quaternary kinematics evolution of the Southern Apennines. Relationships between surface geological features and deep lithospheric structures
Ann. Geofis.
A software package for the adjustment and the analysis of GPS control network
Software Available for Analyzing GPS Deformation
EOS, Transaction, American Geophysical Union, 79, 22, 259 and EOS electronic Supplement
Strain tensor estimation by GPS observations: software and applications
Boll. Geod. Sci. Affini
Active crustal extension in the central Apennines (Italy) inferred from GPS measurements in the interval 1994 and 1999
Geophys. Res. Lett.
Present-day Kinematics of Adria and its Southern Boundary
Eos Trans. AGU
Seismic hazard and seismic zonation of the region affected by the 2002 Molise, Italy, Earthquake
Earthq. Spectra
The October–November 2002 Molise seismic sequence (southern Italy): an expression of Adria intraplate deformation
J. Geol. Soc. (Lond.)
Teleseismic analysis of the 1990 and 1991 earthquakes near Potenza
Ann. Geofis.
Cited by (12)
Rates of geodetic deformation across active faults in southern Italy
2014, TectonophysicsCitation Excerpt :The kinematic parameter estimations were the expected values for the obtained probability distributions and the associated errors were the related standard deviations (Table 3). Within the time period considered in this study, no significant earthquake has occurred in southern Italy nor in the immediate surroundings, apart for the 2002 Molise sequence in the Daunia sector of the eastern domain (Fig. 3), which had a cumulative co-seismic offset of 120 mm for the two main shocks (both with Mw = 5.7) as recorded by GNSS data (Giuliani et al., 2007). Within the western domain, no estimates are available for the offset associated to the 1998 Mw = 5.6 Mercure earthquake.
EMPIRICAL RELATIONSHIP FOR ASSESSING THE NEAR-FIELD HORIZONTAL COSEISMIC DISPLACEMENT USING GPS SEISMOLOGY DATA
2021, Geofisica InternacionalSatellite positioning and geophysics studies in Italy
2015, Rendiconti Lincei