Looking for missing earthquake traces in the Ferrara-Modena plain: an update on historical seismicity

According to the latest issue of the Italian seismic catalogue [Rovida et al. 2011], the area most affected by the May-June 2012 Emilia sequence [for an overview, vide Galli et al. 2012; Tertulliani et al. 2012, this volume] has a centuries-old seismic history of comparatively low magnitude events. These have been on a par with those that have occurred in the nearby Reggiano-Parmense area (where M 4.5 to M 5.5 earthquakes were comparatively frequent in the last century), although less significant than those on record in the portion of the northern Apennines that forms the southern boundary of the Pia-nura Padana (Figure 1), and a lot lower than in the most seismically active zones of the Italian peninsula. Magnitude ca. 5.5 earthquakes are known to have occurred near Ferrara (in 1346, 1561) and in the areas of Finale Emilia-Bondeno (1574, 1908, 1986), Mantua (1901) and Cento (1922). However, this picture might be incomplete, as suggested by the recent discovery of a previously unknown earthquake that occurred in 1639, whose maximum intensity was assessed as 7-8 Mercalli–Cancani–Sieberg (MCS) in Finale Emilia by Camassi et al. [2011a]. [...]

Looking for missing earthquake traces in the Ferrara-Modena plain: an update on historical seismicity et al. 2007], and two similarly-sized earthquakes that occurred in 1796 (M W 5.6) and 1909 (M W 5.5), the latter of which might have had a deep hypocentral location, judging from its wide far field [Meloni and Molin 1987]. To the southwest, in the Carpi-Reggio area, there have been several moderately sized earthquakes, the latest of which was located near Correggio in 1996 (M W 5.4).

Assessing the quality of the information behind the current picture
The studies from which the earthquake parameters of Table 1 were taken by Locati et al. [2011] are of various provenance and of variable thoroughness, and in a few cases they are relatively out of date. Most derive from Guidoboni et al. [2007], a collection that includes both advanced and preliminary studies. The studies derived from Archivio Macrosismico GNDT [1995] and ING [1998] were extremely preliminary ones; the ENEL [1985] studies were at least 27 years old and might be improved by taking into account the latest findings of local historical research (as in the case described by Camassi et al. [2011b]). In a few cases (the earthquakes of 1425,1508,1901), the current catalog inherited the parameters assessed for them by an earlier catalog [Postpischl 1985]; two of these earthquakes have been recently studied by Molin et al. [2008].
Looking at the general picture of local seismicity from a historian's point of view, some peculiarities become obvious. Up to the mid-1800s, for instance, there was a marked tendency for the earthquakes to cluster around Ferrara and Modena, the main towns in the area. A comparison between the seismic histories of Ferrara and Modena and those of lesser towns located midway between them, such as Carpi, Mirandola and Finale Emilia (Figure 2), shows the former to be both longer and more detailed than the latter. This suggests that the original evidence from which these seismic histories were built might have been affected by what the founding fathers Article history Received July 17, 2012;accepted August 23, 2012. Subject classification: Historical seismology, Surveys, measurements, and monitoring, Seismic risk, Seismological data, General or miscellaneous. of modern historical seismology called 'the urban fixation' [Vogt 1986, Vogt and Ambraseys 1991, Vogt 1994; i.e., an inclination to focus on what happened in towns, rather than in villages and rural areas. This was particularly the case in the centuries in which literacy tended to be more widespread among townspeople than in the countryside.
It is also noticeable that earthquakes appear to have been comparatively scarcer in the 1600s to 1700s time window than in earlier centuries. This might reflect a real trend in the seismicity, although it might also be, partly at least, the result of choices made in the selection of historical earthquake data collected by the pre-instrumental tradition of Italian descriptive earthquake studies. This culminated in the great compilation by Baratta [1901], which is responsible, to this day, for the identification of most earthquakes that are included in the current catalog [vide Camassi et al. 2011a]. For a couple of examples, the great 18th century standard collection of medieval chronicles known as Rerum Italicarum Scriptores was undoubtedly instrumental in providing information on many earthquakes observed in Modena and Ferrara before the 1500s. Conversely, the paucity of data available in the 1600s to 1700s time window might be at least partly due to the chance selection of the Bologna gazette as a privileged source of earthquake information by Baratta [1901]. This gazette was one of the oldest Italian periodicals (1678-1796), and it usually published news from Rome and from a few Italian (Venice, Milan, Genoa) and European towns, but not from those at the eastern end of the Pianura Padana, such as Ferrara and Mantua [Camassi and Caracciolo 1994].

A few missing local earthquakes rediscovered
The incompleteness of the current picture of local seismicity is shown by the number of previously unknown, moderately damaging, local earthquakes, for which the information was retrieved by recent studies ( Table 2). Three of the more interesting cases are described below.

1) April 6, 1639 (Finale Emilia, Carpi?)
A recent history of the Jewish community of Finale Emilia [Balboni 2005] quotes an authoritative local history [Frassoni 1778] as having reported an earthquake that occurred on April 6, 1639. This earthquake caused houses and chimneys to collapse in Finale, where a woman was killed. Reliable contemporary confirmation of this scenario can be found in the records of an inspection that was made by the bishop of Modena to the Finale churches in October 1639, which describes the tumbledown conditions of the Chapel of the Most Holy Annunciation "owing to the earthquake" [AAMoN 1639]. A document in the Carpi municipal archives dated as July 3, 1638, and transcribed by D'Orazi [2012] records the damage to the municipal tower of Carpi caused by "the past earthquake". Further research will be devoted to check the date of the Carpi document, and to look out for connections before it and the 1639 Finale Emilia earthquake.
math of the late eighteenth century Italian earthquakes, many such works were produced by supporters and opponents of the then fashionable theory of electricity as the cause of earthquakes [Castelli 2010]. Although this earthquake has been ignored by Italian seismological compila- shows the Virgin and Child with two saints (one of which is Francis Solano, who was venerated as a special protector against earthquakes [Castelli and Camassi 2006]), and the tottering buildings of Mirandola in the background. Further research will be devoted to looking for in-situ evidence of damage (if there is any).

3) May 11, 1778 (Rovereto sulla Secchia, Concordia sulla Secchia, Carpi)
This appears to have been a seismic sequence of some duration that started with a "fearful" shock that was felt in Carpi on May 1, 1778, and continued on to August 25 of the same year [D'Orazi 2012]. The main shock on May 11, 1778, caused irreparable damage to the "most ancient and strong" tower of the Sacchella, located in the nearby hamlet of Rovereto sulla Secchia, which consequently had to be completely demolished. Fearing further collapses, the people of Rovereto and those of nearby Concordia sulla Secchia left their houses and camped in the fields until the end of the shocks. This sequence has also remained so far unknown to any Italian seismological compilation or catalog.

Conclusions
The information available on the historical seismicity of a region subjected to investigation needs to be submitted to critical scrutiny, especially in the aftermath of a strong earthquake, to test our understanding of it, and if necessary, to improve upon it. In this case, the May 2012 sequences can bid fair as the largest known to date for these areas, as the 1570 Ferrara sequence appearently occurred in a more eastward sector of the same seismogenic structure [vide Galli et al. 2012]. Our present knowledge of historical seismicity in the territory that extends from Ferrara to Finale Emilia, and from there to Mirandola and Novi di Modena, appears to be both defective and vague. However, we appear to be well on the way towards an improvement to this situation. Recent studies have retrieved the traces of several damaging earthquakes ( Table 2) that occurred between the 1600s and 1700s that have so far been overlooked by the seismological literature and the parametric catalogs. The likeliest reasons for their fall into oblivion would appear to be that they failed to attract the attention of most outside contemporary witnesses because their highest macroseismic effects occurred in circumscribed, mostly rural, areas, while their attenuation was so strong that they were not felt in any major nearby town (e.g., Bologna, Ferrara, Modena or Mantua). Through the clues retrieved so far, the authors of this paper are about to start an ad-hoc quest for earthquake information in local archives and libraries.