Interseismic Active Deformation in the central-southern Apennine

Abstract

The GPS results are of utmost relevance for the study of the complex plate boundary geodynamics. The lithosphere strain partitioning is part of the seismic cycle. We present the first GPS kinematic pattern obtained during the interseismic phase by a dense episodic GPS network, the Southern Apennine Geodetic Network - SAGNet (Sepe et al., 2009), in the time span 2002-2013. This network is located across the transition zone between central and southern Apennine, including Meta-Mainarde-Venafro and AltoMolise-Sannio-Matese mounts. This region is characterized by seismogenic fault systems responsible, in the past, for several destructive earthquakes of intensity I ≥ IX MCS and, in more recent years, characterised mainly by some moderate magnitude seismic sequences (max magnitude Mw 5.0, December 29 2013) and single small events (Ml < 2.5).SAGNet GPS data were processed by BERNESE sw v.5.0 and the resulting velocities were least-squares combined with the permanent stations velocity field and with the velocity solution of Giuliani et al. 2009. The combined GPS velocity field, shows a perpendicular maximum extension with respect to the Apennine chain of about 2.0 mm/y.The Matese area was hit on December 29, 2013 by a Mw=5.0 (Convertito et al., 2016) earthquake. It was followed by an intense seismic activity until the beginning of February 2014. After the mainshock a GPS survey was carried out on the SAGnet stations. We collected data from 2013, 30 December to 2014, 4 April. The time series of 17 stations are affect by an offsets on the linear drift. The map of horizontal coseismic displacements (Figure 3) shows a sub-radial displacement shape with respect to the epicentre. Larger displacements are observed in correspondence of NE portion of the Matese massif. Considering the Matese Lake Fault as the probable source of the mainshock (dip 65°, strike 116, rake 270 – MLF, Ferranti et al, 2015), we found that the Okada modelling does not fit the observed displacements and only a small fraction of displacements are resolved with a simple slip.Figure 4 resembles the results of previous studies compared with our GPS analysis. We considered seismological analyses, tomographic models, degassing of CO2 data and conceptual model of processes recognized in South Apennine (L. Bisio, et al., 2004; Chiarabba and Chiodini, 2013; Improta et al., 2014; Ventura et al., 2007, R. Di Stefano and M.G. Ciaccio, 2014; Ferranti et al., 2015; Convertito et al., 2016;). The GPS results indicate that the relative motion between Eurasia and Adria plates is responsible of the active deformation in the Apennines. The most important outcomes of this study are: (i) During the interseismic phase the differential motion between Adriatic and Tyrrhenian domains seems to be accommodated in a narrow belt bordering the westward flank of the Sannio Mts, showing a 2 mm/y extension. (ii) The maximum extension does not follow the topographic high of the chain but is shifted toward the eastern outer belt. (iii) No significant GPS deformation is highlighted in correspondence of major and known fault systems where the GPS velocities appear almost steady. We propose that the observed coseismic displacements are only marginally explained by a slip on the MLF fault. The vertical directivity and depth distribution of the seismic sequence (Convertito et al., 2016), the vertical and horizontal heterogeneity of lower crust and upper mantle (Bisio et al., 2004; Di Stefano and Ciaccio, 2014), the high flux of CO2 degassing (Ventura et al., 2007, Chiarabba e Chiodini, 2013 ), the probable presence of pressurized CO2 bodies fed by fluids uprising from the mantle wedge (Improta et al.,2014 ), suggest instead that the seismic sequence could be caused by sub-vertical cracks that originate at the Moho interface and reach the bottom of the seismogenic layer (10km depth).