Two unknown M > 6 historical earthquakes revealed by palaeoseismological and archival researches in eastern Calabria (southern Italy). Seismotectonic implications P. Galli and V. Scionti Civil Protection Department, USSN, Via Vitorchiano 4, 00189 Rome, Italy Introduction The seismicity of the Calabrian pen- insula is one of the strongest of the Mediterranean, both in terms of maxi- mum magnitude and frequency (at least 25 M ‡ 6 earthquakes have oc- curred since 91 BC between the Mes- sina Strait and the Crati basin; Fig. 1). Ninety-five percent of the energy released in the past millennium oc- curred between the catastrophic earth- quakes of March 1638 (Mw ¼ 6.7; Galli and Bosi, 2003) and those of 1905 and 1908 (Ms ¼ 7.5 and Ms ¼ 7.3, respectively; WG-CPTI, 1999). This clustering could be related to the dynamic behaviour of seismo- genetic structures, which triggered each other roughly from the north to the south starting from the 17th cen- tury. Considering that the average time-recurrence for characteristic earthquakes on each fault (1– 1.4 kyr; Galli and Bosi, 2002, 2003) is much longer than the period of completeness of the Calabrian cata- logue (0.5 kyr for M > 6; GdL, 2004), the seismic catalogues missed two ancient earthquakes that struck Calabria. Likewise, they also missed several earthquakes generated by faults that ruptured before the time span covered by written accounts (i.e. silent faults). Recently, there have been investi- gations of unknown historical sources about Calabria (e.g. Camassi and Castelli, 2004), and in particular for the period between 1500 and 1750 (when the amount of the known energy release was less than 20%: Scionti et al., 2006). At the same time, palaeoseismological inves- tigations carried out from Mt. Polli- no to the Gioia Tauro plain began to unravel the tangled skein of the occurrence and recurrence of earth- quakes in Calabria. In this scenario, one of the less investigated areas, both from the historical and seismogenic points of view, is the Ionian side of Calabria. In this paper, we show (i) the palaeoseismological evidence and age of the unknown medieval earth- quake that predates the June 1638 event (Mw ¼ 6.7) along the Lakes fault (Galli and Bosi, 2003; 5 in Fig. 1) and (ii) the existence in archival documents of another un- known strong event (1744, Mw ¼ 6.2) that occurred close to the SE tip of the same fault. Both earthquakes shed light on the poorly constrained seismotectonics of the area. Seismotectonic framework of Calabria The Calabrian Arc is the southern termination of the Apennine thrusts and folds chain. It lies above a NW- dipping subduction slab (Ionian ocea- nic lithosphere; Amato et al., 1993; Piromallo and Morelli, 2003) plunging 500 km into the astenosphere below a backarc basin (S-Tyrrhenian Basin; Gueguen et al., 1997 and references therein). Since lower Pliocene, exten- sion produced N–S and NE–SW trending basins on the inner margin of the Arc (e.g. Crati and Mesima graben, Gioia Tauro basin), all were bounded by primary high-angle nor- mal faults (Tortorici et al., 1995; Galli and Bosi, 2002, 2003; Fig. 1), whereas starting from middle-Pleistocene, Calabria was abruptly uplifted (0.6– 1.3 mm yr )1 -rates, respectively, from north to south; e.g. Dumas and Raffy, 2004; Molin et al., 2004). According to Westaway (1993), the uplift and the extensional faulting in the Calabrian Arc are a result of the isostatic re- bound that occurred when the de- tached deep portion of the Ionian slab sunk, and the non-detached portion rapidly unbent and bounced upwards (see alternative mechanisms quoted in Molin et al., 2004). Almost all M > 6 earthquakes have been generated by the normal faults that drove the opening of the exten- ABSTRACT The seismicity of Calabria is among the strongest in the whole of the Mediterranean, but is not well known on its Ionian side. Palaeoseismological investigations in eastern Calabria and archival research have tried to shed light on earthquake recurrence and seismogenetic source geometry. Trench inves- tigation across the Lakes fault (Sila massif) shows this fault to be definitively responsible for the Mw ¼ 6.7, 1638 earthquake, and provides a well-constrained age range for its penultimate event (early 7th century AD). On the other hand, after research in local and national archives, another previously unknown relevant historical earthquake has been discovered (1744, Mw ¼ 6.2), which struck an area conterminous to the SE tip of the Lakes fault. We hypothesize a NW–SE active Sila fault system that accommodates regional N–S extension through normal-to-sinistral segments, each one capable of generating earthquakes with magnitude in the range of 6.2–6.7. Since 1638, this system ruptured almost entirely from NW towards SE (1638–1744–1832), with the exception of the northwestern segment, the ‘silent’ Cecita fault. Terra Nova, 18, 44-49, 2006 Correspondence: Dr Paolo Galli, Diparti- mento Protezione Civile, Servizio Sismico Nazionale, Via Vitorchiano 4, 00189 Rome, Italy. Tel.: 00 39 06 6820 4892; fax: 00 39 06 6820 2877; e-mail: paolo.galli@ protezionecivile.it Ó 2006 Blackwell Publishing Ltd 44 doi: 10.1111/j.1365-3121.2005.00658.x