Historical earthquakes and damage patterns in Potenza ( Basilicata , Southern Italy )

This paper analyzes three historical seismic damage patterns in Potenza, the main town of the Basilicata Region, 
in Southern Italy. We refer to the 1826, 1857 and 1930 earthquakes, which hit the town with intensities ranging 
from VI-VII to VIII-IX MCS. In order to depict the seismic effects, we analyzed original documents, most of 
them never consulted before. The events have been located on an urban map of Potenza dating back to 1875. 
This research represents the starting point for further investigations, with the goal of highlighting the causes of 
the anomalies in the distribution of the effects.


Introduction
The use of historical data to analyze the effects of past earthquakes can meaningfully increase the knowledge about the seismic behaviour of the building-soil system.This approach can highlight urban areas where anomalously large damage occurred in the past.Further investigations may be able to determine the causes of the differential effects, which are usually due to geological/anthropic features of the territory and/ or building vulnerability (e.g., Guidoboni et al., 2003;Gizzi, 2006).In this way, potentially 'weak' urban zones can be identified so that suitable prevention strategies can be implemented.
To this purpose, we analyzed macroseismic data related to three historical earthquakes in Potenza, the main town of the Basilicata Region, in Southern Italy.Our findings constitute the starting point for further analyses aimed at highlighting the causes of 'anomalies' in damage distribution.In detail our research will make it possible to determine the areas subjected to heavier damage during historical events, which could be at risk in the event of future earthquakes.Further research based on data collection and analysis (e.g., field surveys, boreholes and geophysical prospecting) may be able to determine the impact of geological, geomorphological, geotechnical and hydrogeological features of the site on damage distribution.In this way, important implications will be gathered for the seismic microzonation of Po-tenza, based not only on the traditional approach but also on damage already experienced.
We studied the most damaging historical earthquakes (in 1826, 1857 and 1930) for which it has been possible to map the affected buildings.These events struck the town with intensities ranging from VI-VII to VIII-IX MCS (Boschi et al., 2000).We analyzed documents preserved in Potenza and in the Naples State Archive as well as in the Potenza Municipal Archive and in the Potenza Monuments and Fine Arts Office

Historical earthquakes and damage patterns in Potenza
(Basilicata, Southern Italy) Archive.These data, in conjunction with urban, technical and cadastral information, enabled us to describe the three urban damage patterns.

Outline of the seismic history of Potenza
Potenza, a town with a population of about 70000 (Istat, 2005), is located on a hill in the Southern Apennines where an axial-active, 30 to 50 km wide seismic belt produces strong earthquakes (Boschi et al., 1999).Large past earthquakes originated in both inner areas of Basilicata and in areas located at the western and north-western boundaries, between Basilicata and Campania (fig.1).
In historical times, Potenza was affected by at least five earthquakes with intensities higher than or equal to VIII MCS (table I).
The first documented event dates to 1273 (Io=VIII-IX MCS, Me=5.7).The only available written account states that in Potenza a large number of buildings suffered heavy damage.According to these data, the local effects were VIII MCS (Boschi et al., 2000).
About three centuries later, another seismic event occurred along the western boundary between Campania and Basilicata.The two main shocks (on July and August 1561) caused considerable damage, mainly in Potenza and Salerno.The first events (Io=IX-X MCS, Me=6.5) caused the collapse of 10 houses in Potenza, with no casualties (Boschi et al., 2000).
About a century later, the 1694 Irpinia-Basilicata earthquake (Io=X MCS, Me=6.8) caused the total destruction of 14 towns and great damage in 18 towns and villages along the Apennines.Pacichelli (1695) stated that: «In Potenza several buildings, the Church and the Trinità Tower collapsed.Five people died …» («Cadde in Potenza con più fabbriche, la chiesa e la Torre della SS.Trinità, con la morte di cinque cittadini …»).
The seismological importance of this earthquake led the Irish engineer Robert Mallet to visit the affected towns, including Potenza.For this locality, the scientist noticed that «…The damage done was very diffused over every part of the town, though greatest upon the free lying flanks, to the east, west, and N.W., and many buildings are prostrated or fractured …» (Mallet, 1862) (fig.2).
The last historical earthquake that affected the town happened on 23 July 1930.The Irpinia quake had an epicentral intensity and magnitude of Io=X MCS and Me=6.7 respectively, and hit at 00:08 GMT an area of about 6500 km 2 , killing 1404 people (Boschi et al., 2000).
In several towns at the border between Campania and Basilicata, everything turned to dust and rubbish.The most affected area was the territory ranging from Melfi to Ariano Irpino.In Potenza, the damage was serious: 80 buildings were damaged (Alfano, 1931).
About thirty years ago, the Italy's largest natural catastrophe since World War II occurred.We refer to the 1980 Irpinia event (Io=X MCS, Me= =6.7).This earthquake devastated a wide area of the Southern Apennines, especially Campania and Basilicata where about thirty towns were completely destroyed.In Potenza 10 000 people became homeless and public buildings suffered heavy damage (Il Mattino, 08/12/1980; Il Nuovo Corso 10/02/1981; Gazzetta del Mezzogiorno 30/12/1980).
Ten years later, on May 5, 1990, the Potenza seismic sequence occurred along an E-W striking vertical plane (Demanet et al., 1998).The local shaking felt was VI MCS, and the maximum intensity was recorded in Pietragalla (VII MCS), about 15 km from Potenza (Tertulliani et al., 1992).

Seismic damage patterns: methodology and results
As stated previously, this analysis focuses on the urban damage patterns from the earthquakes of 1826, 1857 and 1930.Published written accounts were not sufficient to locate urban effects, mainly for two reasons.First, there is little detailed information about damage; the second reason is the lack of 'toponymy markers' suitable to locate damage episodes on a map.
To overcome this lack of information, we analyzed unpublished written sources from the Potenza and Naples State Archives, as well as the Potenza Municipal Archive and the Potenza Monuments and Fine Arts Office Archive.
Bibliographic research also was undertaken to gain information about the historic, urban and social background of the town.Understanding these aspects allowed us to better interpret the damage as quoted in the written sources.All damage episodes were located on an urban map from 1875, approximately halfway in time between the first and last earthquakes.
The seismic effects have been classified according to the EMS-98 scale (Grünthal, 1998), which considers the following five damage classes for masonry buildings: slight, moderate, substantial to heavy, very heavy and destruction.
In order to correctly extract information from the historical accounts, we had to reconstruct administrative actions taken after each earthquake; a brief summary follows.
After the 1826 seismic event, a central commission was established with the task of supervising the rebuilding of damaged houses by distributing money to the owners of those houses.In Potenza, a local commission surveyed the damaged buildings as well, to decide whether or not to demolish severely damaged houses.These commissions were composed of administrators, technicians and representatives of the town clergy.
In order to detect damage, we first analyzed the resolutions of the local commission and correspondence between the municipal commission and the Intendenza (prefecture).We also studied about 20 letters mailed by house owners to the central commission so as to obtain subsidies to repair or rebuild the dwellings.Much of this correspondence contains descriptions of seismic effects; however, the owners tended to overestimate damage.Letters that cite unambiguous damage have been considered reliable (e.g., collapse of a wall, serious damage linked to demolition imposed by the municipal commission, etc.).The owners' damage descriptions were correlated with data drawn from the documents of the commissions.Therefore, by critically analyzing these written sources, we obtained more information about damage in urban areas.Several sources analysed are documents unpublished elsewhere (see Appendix).
In order to locate damaged buildings on the urban map, we had to address other problems.The locations (road, square, etc.) of houses are commonly not cited in historical records, nor are they referred to in old toponymy.To overcome these difficulties, we conducted a crosscorrelated analysis of cadastral data, old and current toponymy and site surveys.The descriptive French cadastre (so-called Catasto provvisorio) that dates back to the first decades of the nineteenth century was consulted.This guaranteed a nearly one-to-one correspondence with locations of buildings at the time of the earthquake.
Once the correct location was identified, the affected houses were located on the urban map, which dates to the second half of the nineteenth century.To obtain a better match with the urban setting at the time of the earthquake, urban features gathered from bibliographic analysis were integrated with the urban map (e.g., Buccaro, 1997).
The analysis of this earthquake highlights effects on the western flank of the hill and in the central-western part of the historical centre (fig.3).
However, available written sources cannot completely depict the event.An Even though this research cannot indicate the causes of differential effects, it seems appropriate to emphasize the vulnerability of the buildings.The commission entrusted with surveying the damaged houses stressed that dwellings suffering damage were old, with preexisting cracks (Sannino, 1990).Moreover, after the earthquake, no work was undertaken toward systematically strengthening the buildings, due in part to the house owners' poverty.Thus, the stronger 1857 earthquake happened at a time of structural vulnerability.
Following the 1857 earthquake, central and local commissions were adopted as well.The damage pattern is here mainly 'reconstructed' on the basis of unpublished documents (see Appendix).
The resolutions of the commissions regarding the demolition of severely damaged houses and the distribution of subsidies to rebuild or to repair the houses are all unpublished.Also unpublished is the correspondence between house owners and the central commission intended to obtain subsi-    (Mallet, 1862, stereoscopic photo no. 306).Huts are also evident in the foreground.dies.The subsidies were assigned if a twofold condition was met: significant damage to the buildings and poor house owners.Obviously, this criterion filters damage evaluation and must be considered in the analysis of the effects (fig.4).
We also analyzed the records of the Consiglio Edilizio (Angelini, 1995), a technical body established for 'urban planning'; this was very important in mapping the seismic effects.This source provided data about affected houses not included in other analyzed documentation, owing to the previously cited economicadministrative filter.
More data were gathered by cross-correlation of these last-mentioned dossiers and the documentation produced by the local and/or central commissions.In this way, information was obtained on both the damage and the locations of buildings.For locating dwellings, we discovered additional information from other kinds of sources coeval to the earthquake (e.g., list of house owners who made use of public works or list of dwellings demolished in order to construct new public buildings (ASCPz, Elenco complateari …, 1857; ASCPz, Elenco delle case demolite …, 1857).The French cadastre was not useful for locating damaged buildings because it was not coeval with any earthquake.
Damage to public or religious buildings was considered in both the archive sources and the surveys or photos attached to the Mallet report (Mallet, 1862) (figs. 5, 6).These photos helped locate some collapsed or affected buildings on the urban map.
At least ten aftershocks were felt in Potenza, lasting until November 1858 (Rossi, 1868).Therefore, it is clear that the classified damage could include cumulative effects of the earthquakes.
Using all the sources mentioned above, the localised damage events were pinpointed on the 1875 urban map.Figures 7 and 8 indicate that the effects were mainly localized in the western and central-western sectors of the historical   centre, the so-called Portasalza and Portamendola quarters, where many buildings collapsed.The eastern flank of the hill on which the town lies was also heavily hit.
For the 1930 earthquake, the damage pattern was also 'reconstructed' using original sources that have never been referenced elsewhere.We refer to the written sources from the Potenza Municipal Archive and the Potenza Monuments and Fine Arts Office Archive (see Appendix).
The Municipal Archive sources include correspondence between the owners and the prefecture.We also examined two lists of damaged buildings prepared by the Office of Civil Engineers (Genio Civile) and the town council.These data are the only available documentation, because Potenza was not included among localities where a public subsidy was granted for rebuilding.Because there was greater damage to other towns (e.g., Gizzi andMasini, 2004, 2006), detailed damage to Potenza was not documented by the Office of Civil Engineers.
The available information refers to damage to about 90 buildings, including public and religious buildings.These data almost agree with Alfano (1931), who reported damage to 80 buildings.This scientist probably only considered dwellings in his computation.
The Monuments and Fine Arts Office Archive was useful in gathering information about damage to two churches.
Due to the documents consulted, accurate damage classification was not possible in many cases.However, completing the damage map represents a significant increase in knowledge about seismic impact at this location.Indeed, the 'scenario' map clearly shows that damage was mainly localized in the central-western and western parts of the historical centre (fig.9).

Conclusions
This paper sheds light on the effects produced to the town of Potenza (Basilicata Region, Southern Italy) by three historical earthquakes: those of 1826, 1857 and 1930.This was mainly achieved by analyzing original written documents, often never consulted before.This study produced three urban maps showing the houses and other buildings that sustained damage.An analysis of the three damage patterns shows that the western flank of the hill and the central-western part of the historical centre suffered the most severe damage.
Even though it is still not possible to define the causes of these differential effects, the data would suggest an influence of building vulnerability, especially for the 1826 seismic event.
However, further research, based on data collection and analysis (e.g., field surveys, boreholes and geophysical prospecting) would clarify the roles of geological, geomorphological, geotechnical and hydrogeological features of the site as regards damage patterns.This would produce important implications for the seismic microzonation of Potenza, based on damage already experienced.

Fig. 1 .
Fig. 1.Strongest historical earthquakes that affected Potenza.The circle diameter is proportional to the epicentral intensity (MCS).The black circles show the seismic events analysed in this paper.Data are drawn from Boschi et al. (2000).
engineer entrusted to verify the effects in Potenza wrote that: «… mostly damaged buildings are the Cathedral, the Church of San Gerardo and its bell-tower, the Church of Santa Trinità, the Palace of Intendenza … and several dwellings of rich and poor owners whose list is long …» («Gli edifici che più han sofferto sono l'Episcopio, la Chiesa ed il Campanile di S. Gerardo, la torre di San Carlo, la chiesa della SS.Trinità, il convento dei Riformati sotto il titolo di S. Maria, il Palazzo d'Intendenza e molte abitazioni di ricchi e poveri, di cui sarebbe lungo tesserne l'elenco …») (ASNa, Relazione dell'Ingegnere provinciale Marino Massari al Signor Intendente …, 1826).

Fig. 3 .
Fig. 3. 1826 damage 'scenario' for Potenza.The map highlights effects on the western flank of the hill and in the central-western area of the historical centre.

Fig. 4 .
Fig. 4. Number of damage events pinpointed on the urban maps.Events are plotted as two different group of EMS damage classes (D1-D3 and D4-D5).The predominance of the higher classes of damage is due to the economic-administrative filter (see text).

Fig. 5 .
Fig. 5. Damage to the Palace of Intendenza (on the top in the right corner) according to the Mallet's report(Mallet, 1862, stereoscopic photo no.306).Huts are also evident in the foreground.

Fig. 7 .
Fig. 7. 1857 damage 'scenario' for Potenza.It is evident that the effects are mainly localised in the western and central-western sector of the historical centre, the so-called Portasalza and Portamendola quarters, where many buildings collapsed (see also fig.8).

Fig. 9 .
Fig. 9. 1930 damage 'scenario' for Potenza.It can be noticed that the damage is especially located in the central-western and western portion of the historical centre.