First evidences of fast creeping on a long-lasting quiescent earthquake normal-fault in the Mediterranean

Abstract

S U M M A R YA key issue in our understanding of the earthquake cycle and seismic hazard is the behaviour of an active fault during the interseismic phase. Locked and creeping faults represent two end-members of mechanical behaviours that are given two extreme rupturing hazard levels, that is, high and low, respectively. Geophysical and space geodetic analyses are carried out over the Pollino Range, an extensional environment within the Africa-Eurasia plate boundary, to disclose the behaviour of the long-lasting quiescent Castrovillari normal fault. Fault trenching evidenced at least four large earthquakes (6.5-7.0 M w ) in the past and an elapsed time of 1200 yr since the last event. Inversion of Differential Interferometric Synthetic Aperture Radar and Global Positioning System over a decade shows fast creeping at all depths of the fault plane. The velocity-strengthening creeping zone reaches maximum rates 20 mm yr −1 against an average rate of about 3-9 mm yr −1 . It limits the southern-weakening locked part of the fault. An essential condition for the generation of a large earthquake on the Castrovillari fault, as has occurred in the past, is a rupture through the velocity-strengthening zone. The Castrovillari fault yields the best evidence for being both a strong and weak fault during its earthquake cycle. Creeping at rates faster than its tectonically driven ones, it must thus consist of a mix of unstable and conditionally stable patches ready to sustain a sizeable earthquake. Quantifying and mapping the slip rate over the fault plane is important because they influence fault moment budget estimate and helps to constrain constitutive laws of fault zones. Aseismic slip also redistributes stress in the crust, thereby affecting the locations of future earthquakes.

Aseismic transient driving the swarm-like seismic sequence in the Pollino range, Southern Italy

“…The centimetre scale creeping inferred by Sabadini et al (2009) is not consistent, if persisting over time, with the lack of surface expression of high deformation along the trace of the Castrovillari fault. However, we cannot rule out a largely aseismic transient slip in the Mercure basin as the initiating and/or driving mechanism of the swarm sequence.…”

“…Creeping was invoked for the Castrovillari fault by Sabadini et al (2009) on the basis of InSAR data and local GPS campaigns before the onset of the earthquake swarms. The centimetre scale creeping inferred by Sabadini et al (2009) is not consistent, if persisting over time, with the lack of surface expression of high deformation along the trace of the Castrovillari fault.…”

From Geodetic Imaging of Seismic and Aseismic Fault Slip to Dynamic Modeling of the Seismic Cycle

“…A few examples of continental faults creeping at the surface have long been known, such as the creeping segment of the San Andreas Fault in California (de Michele et al 2011, Titus et al 2005, Tocher 1960, the Hayward Fault in California (Bonilla 1966, Cluff & Steinbrugge 1966, and the North Anatolian Fault in Turkey (Ambraseys 1970, Cakir et al 2005, Kaneko et al 2013. In addition to the LVF, reviewed above, examples of faults creeping at shallow depth have now accumulated thanks to GPS geodesy and InSAR (e.g., Donnellan et al 2014;Doubre & Peltzer 2007;Duquesnoy et al 1994;Hreinsdottir & Bennett 2009;Huang et al 2009;Jolivet et al 2012Jolivet et al , 2013Kaneko et al 2013;Lohman & McGuire 2007;Sabadini et al 2009;Shirzaei et al 2013;Sylvester et al 1993;Szeliga et al 2012;Wei et al 2011). Creep seems to be common on oceanic transform faults and normal faults at oceanic spreading centers: Bird & Kagan (2004) found Z s to be <1 km on these types of faults, and McGuire et al (2005) found that 85% of slip on transform faults along the East Pacific rise was aseismic (McGuire et al 2005).…”

Plate Deformation

“…After the analysis shown in Section 6.04.6.1, which appeared in the international literature in 2009 (Sabadini et al, 2009), the Pollino seismogenic zone, indicated by the square in Figure 52(c), has been struck by a normal fault Mw ¼ 5.2 earthquake on 26 October 2012, epicenter in Mormanno, a locality around 10 km roughly northwest with respect to the site of Morano (the mora (Morano) site of Figures 52-56). This earthquake did not occurred in the southern part of the Castrovillari fault of the previous figures, which, according to our results, was already prone to generate a sizable earthquake, which means magnitude of Mw ¼ 5.0 at the least, in the white part of the fault plane of Figures 55 and 56: although the epicenter of this quake was northwest with respect to the locked part of the Castrovillari fault as envisaged by our findings, the 50 nanostrain year À1 measured in the area (Figure 52(a)) suggests that this large extensional activity is more widespread with respect to the SISMA surveyed area, so that the results of this section represent an underestimate, in space, of the seismic hazard of the whole Pollino seismogenic zone.…”

“…Roughness is 1 mm km À1 , the offset is À2.7 mm year À1 , and other parameters can be found inSabadini et al (2009), from which the figure is redrawn. At the bottom panel, DInSAR data including an offset (left-hand panel), model results (center panel), and misfit (right-hand panel).…”