Istituto di Fisica, Università di Urbino, Urbino (PU), Italia

In December 1456-January 1457 a major earthquake sequence took place across the central and southern Apennines (southern Italy, Calabrian Arc excluded), including southeastern Apulia. A recent re-evaluation of the (a) revised damage pattern for this multiple earthquake, (b) deeper seismicity of the southern Apennines – Adriatic foreland interface and (c) deep-seated regional E-W structures, led to the identification of at least four seismogenic sources, responsible for the main sub-events of the multiple 1456 earthquake. Based on various seismological, macroseismic and tectonic constraints, these causative faults are thought to exhibit an oblique right-lateral motion along fault segments roughly E-W oriented. Such segments are portions of well-known inherited regional E-W trending shear zones (like the Molise-Gondola shear zone), at various latitudes between (from north to south) the Maiella Mts. and the Vulture volcanic complex. This system would therefore imply the cascade reactivation of such shear zones favorably oriented with respect to present stress field, with a transtensional mechanism. More than one catastrophic historical earthquake that occurred in southern Italy suggests the nearly simultaneous activation of multiple sources across widely spaced (+/- 30 km) portions of independent E-W faults. Being the strongest (by magnitude and damage area) among these major earthquakes, the 1456 sequence can be considered as a template for such mechanism of multiple activation of distant sources yet within a short time window. This hypothesis invokes a possible stress interaction between multiple sources falling within neighboring domains. We investigated Coulomb stress changes related to the main sub-events of the multiple 1456 earthquake to analyze fault interaction and stress transfer mechanisms. An evident positive correlation between the calculated Coulomb stress increase and two major seismogenic sources is found. Therefore, the spatial redistribution and enhancement of static stress caused by the stronger events may promote rupture on adjacent faults that are close to the failure threshold. A more general case may be considered imposing a pre-existing stress field or assuming different values for the friction coefficient. To the extents of present knowledge and investigation, these E-W trending earthquake sources are active between ca. 10 and 20 km at depth in the sector of the southern Apennines east of the chain axis, that is to say in the seismogenic macroregion bounded by the thrustbelt (to the west) and by the Apulian foreland (to the east). The stress patterns caused by these faults are consistent with the large NW-SE trending pure extensional sources found along the southern Apennines axis.

On July 23rd 1930, a strong earthquake (Ms=6.6) occurred in the Irpinia region, the most seismically active area of the Southern Apennines (Italy). Destructive effects were reported in a wide area of about 6300 km2, causing more than 1400 victims. The same region had already been struck by several large earthquakes in 1456 (Me=6.9), 1694 (Me=6.0), 1702 (Me=6.0), 1732 (Me=6.6), and 1910 (Me=5.9). Other major events have hit Irpinia since the 1930 earthquake, including that of 1962 (Mw = 6.2) and the catastrophic one of 23 Nov 1980 (Mw = 6.9). Formerly published studies concerning the 1930 Irpinia event include analysis of macroseismic data, first motion polarities and a single station waveforms. By using the available bulletins and the historical seismograms, in our previous study we estimated the source parameters in terms of focal mechanism, magnitude, hypocentral location and seismic moment. Fault length, rupture velocity and other characteristics are also obtained by performing body waveform inversion for moment rate retrieval. These results are here used to study the static stress transfer between the 1930 Irpinia earthquake and subsequent large events like the 1962, and 1980 ones in order to investigate the possible fault interaction and earthquake triggering. To improve our knowledge on the region of the1930 event, we also study the Coulomb stress field related to E-W trending seismogenic sources, responsible for the main sub-events of the multiple 1456 historical earthquake. Modelling of such effects is useful both to obtain more information on seismogenic sources and to gain an improved evaluation of seismic hazard in this region.

In this work we analyse the most important earthquakes of the 20th century occurred in the Altotiberina Valley in 1917, 1918, 1919 and 1948; in particular instrumental relocation and Ms magnitude estimation are re-evaluated. The area investigated, the Sansepolcro basin, is characterized by the presence of important earthquakes in the past with estimated intensity even larger than IX MCS (the 1352 Monterchi earthquake, the 1389 Boccaserriola, the 1458 Citta' di Castello, the 1781 Cagliese and the 1917 Monterchi-Citerna earthquakes, CPTI Working Group, 2004) and by a surprisingly scarce instrumental seismicity compared to the adjacent areas struck by high seismicity (Castello et al., 2005; Ciaccio et al., 2006). In particular, the area north of Sansepolcro has been struck in recent years by four minor sequences, occurred between 1987 and 2001 with magnitude ranging from Ml3.0 to Mw4.7. The relocation of these earthquakes is particularly critical and in an important issue. An instrumental and precise location is critical for the complexity of the problems associated with the study of seismograms prior to the rst half of the twentieth century and is relevant because in the surrounding regions higher seismicity is observed. Regarding this peculiarity of the area, it’s very important to detect the localization of the historical earthquakes: in particular, the 1917 event is often associated to the possibility that the regional low angle Altotiberina Fault (Barchi et al., 1998) is able or not to nucleate large- or moderate-magnitude events, being historically located close to its surface (Boncio and Lavecchia, 2000).

The Southern Apennines chain is related to the west-dipping subduction of the Apulian lithosphere. The strongest seismic events mostly occurred in correspondence of the chain axis along normal NW–SE striking faults parallel to the chain axis. These structures are related to mantle wedge upwelling beneath the chain. In the foreland, faulting develops along E–W strike-slip to oblique-slip faults related to the roll-back of the foreland. Similarly to other historical events in Southern Apennines, the I0 = XI (MCS intensity scale) 23 July 1930 earthquake occurred between the chain axis and the thrust front without surface faulting. This event produced more than 1400 casualties and extensive damage elongated approximately E-W. The analysis of the historical waveforms provides the chance to study the fault geometry of this ‘‘anomalous’’ event and allow us to clarify its geodynamic significance. Our results indicate that the MS = 6.6 1930 event nucleated at 14.6 ± 3.06 km depth and ruptured a north dipping, N100 E striking plane with an oblique motion. The fault propagated along the fault strike 32 km to the east at about 2 km/s. The eastern fault tip is located in proximity of the Vulture volcano. The 1930 hypocenter, similarly to the 1990 (MW = 5.8) Southern Apennines event, is within the Mesozoic carbonates of the Apulian foredeep and the rupture developed along a ‘‘blind’’ fault. The 1930 fault kinematics significantly differs from that typical of large Southern Apennines earthquakes, which occur in a distinct seismotectonic domain on late Pleistocene to Holocene outcropping faults. These results stress the role played by pre-existing, ‘‘blind’’ faults in the Apennines subduction setting