Gianluca Valensise, of the Earthquake Department of INGV, Rome, is a seismologist with a geological background, an INGV research manager, and the author of numerous studies on active faults in Italy and other countries. In particular he is the “founder” of Italy’s Database of Individual Seismogenic Sources (DISS, http://diss.rm.ingv.it/diss/). He has spent over 30 years of his career exploring the relationships between active tectonics and historical seismicity, with the goal of merging geological observations with the available evidence on the largest earthquakes of the past.
Recently, with other colleagues, he published a work that proposes a sort of vulnerability ranking of Apennines municipalities. We discuss it below.
Luca, you are an earthquake geologist. You deal with active faults, seismogenic sources, past earthquakes, seismic hazard. Recently, with other colleagues, you have ventured into the theme of seismic vulnerability of the Italian building heritage. How come this choice?
First of all let me recall that I am a researcher, but also a citizen who is in a position to be able – and having to – to do something immediately useful for his own country.
That said, I believe that it all started ten years ago, following the 2009 L’Aquila earthquake. That the extreme vulnerability of the built-up area was one of the causes of the disastrous results of some earthquakes had already become clear to me from before, if only for having studied the effects of the strongest Italian earthquakes of the ‘900; from that of 1908 in the Strait of Messina to that of Irpinia in 1980, passing through the area of Fucino, devastated by the earthquake of 1915.
But the story actually begins even earlier, with the earthquake of San Giuliano di Puglia in 2002. That earthquake showed everyone that some of the choices made in the defense – or lack of defense – from earthquakes can be so disastrous as to frustrate both the material culture accumulated by those who live in the seismic zones, both the technological advancements in the construction industry: these advancements concern not only reinforced concrete structures, but also those built with load-bearing masonry.
The 2009 L’Aquila earthquake showed almost a complete reversal of the “normal” situation: excluding cultural assets, for which other rules apply, the maximum number of collapses and victims was recorded in post-WWII buildings, while those of previous eras – including the eighteenth-century buildings of the historic city – overall responded quite satisfactorily (this different performance also includes the choice of the site, sometimes disastrous in the case of some recent buildings).
There is little to add on the case of the May 2012 earthquakes in Emilia. The damage was dominated by the collapses in ecclesiastical architecture – to a large extent inevitable – and in the industrial one, which instead represented a bolt from the blue for everyone. I would like to remind you that the collapsed warehouses had been built without considering the possibility of significant lateral seismic actions, which have therefore played a major role in causing collapses that are apparently disproportionate to the severity of the earthquake itself. The previous antiseismic code did not require buildings in that area to be earthquake-proof: nevertheless, this remains a tragedy in the tragedy, if we reflect on the enormous disproportion between how little it would have costed to substantially reduce the vulnerability of those industrial premises (provided, however, that the reinforcements had been planned ahead of construction), and the price paid by those communities in terms of human lives and damage to the local (and national) economy.
Finally, there is the case of the 2016, Central Italy earthquakes, and of the dualism between the historical center of Amatrice, that practically disappeared from the map, and that of Norcia, that has entered a path of rebirth, although this path is fraught with many difficulties. This dualism has spurred our research (1, 2).
The basic thesis that you support is that – perhaps I simplify – the seismic vulnerability of settlements increases with the temporal distance from the last destructive earthquake. In a sense, you argue, after a destructive earthquake, repairs and reconstructions are carried out which reduce the overall vulnerability; then the memory of the event fades out and the vulnerability increases. Is this so?
The thesis in extreme synthesis is the one you have outlined, but I must make two premises. The first concerns the data used: in order to rely on homogeneous and good quality data, we have chosen to analyze only peninsular Italy, and in particular the Apennines chain. The second one has instead a methodological character: in our work we use jointly the historical observations, in the form of the seismic history of each single municipality, and the geological observations, which within a vast territory allow us to isolate those municipalities that are directly located above the great seismogenic sources. As such these municipalities will experience strong ground shaking, sooner or later (see the image below). It should be noted that the historical data would in many cases allow to go even below the municipal scale, but in order to grant a homogeneous representation and to be able to relate to the ISTAT data we have decided to bring everything back to the single municipality.
Composite Seismogenic Sources taken from the DISS database (DISS Working Group, 2018: http://diss.rm.ingv.it/diss/) and the strongest earthquakes (Mw 5.8 and larger) in the CFTI5Med catalog (Catalog of Strong Earthquakes in Italy, Guidoboni et al., 2018). Each source represents the surface projection of the fault at seismogenic depth. The sources in yellow outline the system of large extensional faults running along the crest of the Apennines and have been used for this research. Each source is surrounded by a 5 km buffer whose role is to take into account the uncertainties inherent in its exact location, and therefore its exact distance from the inhabited centers that surround it or lie above it (from Valensise et al., 2017: see Note 1).
It is from the municipalities so selected (see the figure above) – 716 for the whole of central and southern Italy, from Tuscany to Calabria – that we then ranked the attitude of each community to underestimate the level of local danger, and therefore to fatally lower the guard on the issue of building vulnerability. It would have been useless to consider all municipalities, including those that lie far from the large seismogenic sources, because one thing is a dilapidated building in a scarcely seismic area, such as most of the Tyrrhenian side of the Apennines, another is if that same building is located in Amatrice. We wanted to elaborate a “ranking” of the vulnerability forgotten by citizens and their administrators, and we have put in place the best geological, geodynamic and historical knowledge available today – an almost unique heritage in the world – to achieve this goal. One last observation: the data we used are frozen at pre-2016, so our ranking does not take into account the latest earthquakes in the central Apennines.
In this way you have drawn up a sort of seismic vulnerability ranking of the Apennine settlements, based essentially on the temporal distance from the last destructive event. Can you illustrate this ranking a little?
We have ordered our 716 locations (see figure below) as a function of distance over time since the last VIII intensity shaking (Mercalli-Cancani-Sieberg or MCS scale:): a level of intensity that we believe marks a boundary between the simple repair of old buildings and the need to demolish and rebuild them from scratch, with a presumably drastic reduction in vulnerability.
Distribution of the 716 municipalities (representative of the entire municipal areas) selected with the procedure described in the text (from Valensise et al., 2017). The areas outlined in yellow represent the surface projection of the large seismogenic sources that run along the crest of the Apennines. The map shows:
– in purple: 38 municipalities for which historical sources report only minor damage;
– in red: 315 municipalities that in our ranking correspond to the municipal areas that have not suffered destructive earthquakes since 1861 (the year of unification of Italy);
– in black: 363 municipalities ordered according to the distance in time from the latest destructive earthquake, which occurred after 1861.
The reference to 1861 is purely conventional. The year 1861 represents a historical watershed that is also essential for earthquakes, with variable effects on a case-by-case basis (just think of the very effective Bourbon seismic regulations, that were abolished following the Unity of Italy).
The first 38 localities are those that have never experienced a shaking of the set level: following are those where that level was reached or exceeded many centuries ago, while in the end we find the places that have suffered for the most recent earthquakes, and therefore have been presumably reconstructed with anti-seismic systems.
Our elaborations are easily accessible to anyone through a dedicated website, which shows our ranking both in table and on map, and allows to explore the seismic history of each municipality (3). The only other parameters we show, without using them for the moment, are the resident population and the percentage of pre-1918 buildings, both from ISTAT data.
To illustrate the implications of our study I will give examples taken from the ranking itself. A striking case is that of the Mid-Serchio Valley, with several localities in the highest part of the ranking, that collects the municipalities that have never experienced a VIII degree in history: going from NW to SE we find Gallicano (193°), Coreglia Antelminelli (192 °), Borgo a Mozzano (31°), at Bagni di Lucca (32 °), all centers around 4,000 to 6,000 inhabitants, all in the province of Lucca. Only Barga , the pleasant mountain village celebrated by Giovanni Pascoli that is also the main center of the area, is presumably safe (595°).
It is easy to see that the position in the ranking goes up – thus worsening – moving towards the SE, i.e. moving away from the source of the 1920 earthquake in Garfagnana, also known as upper Serchio Valley. There is no doubt that the two portions of the valley are similar, but the seismotectonic data suggests that while the northern part suffered its “great” earthquake less than a century ago, the fault beneath the southern part is historically silent. According to the CFTI5Med catalogue, in 1920 Barga suffered a VIII degree, and the earthquake “… damaged 75% of the buildings, mostly inhabited by a poor population, causing the total collapse of many houses …”. Will this reconstruction suffice to save Barga from the next strong earthquake in the Serchio Valley? Things will probably be better than in the most downstream municipalities, also because, if it is true that according to ISTAT, 37% of Barga’s housing stock is pre-1918, i.e. more than a century old, this share of buildings is probably made up of houses that resisted the 1920 earthquake: either because they were built better, or because they were built where the seismic response was less severe than the average, or because of a combination of these two circumstances.
Another example I would like to take concerns the Calabrian-Lucanian border, between the provinces of Potenza and Cosenza in southern Italy. The case is similar to the previous one, but definitely more evident. We are in fact in one of the few portions of the Apennines chain that have never suffered a strong earthquake in historical times, even if the completeness of the seismic record of the area does not exceed a few centuries (with Emanuela Guidoboni in 2000 we wrote a small contribution precisely on this theme: see Note 4). The area had already been identified as a possible “seismic gap” by Japanese seismologist Fusakichi Omori within a study he conducted on the largest earthquakes of the Italian peninsula. In the area in question lie Mormanno (Cosenza, 29°) and Rotonda (Potenza, 30 °), never affected by a strong earthquake, but also Viggianello (Potenza, 178°), hit by a VIII-IX degree in the earthquake of January 26 1708 – which according to the CFTI5Med catalogue “… seriously damaged the village causing extensive destruction and numerous victims …”. On 25 October 2012 this area was hit by an earthquake with Mw 5.3, which tested the solidity of the buildings but above all it spurred a vast effort for the reduction of building vulnerability: a very local circumstance, linked to the occurrence of an earthquake that is not destructive but sufficient to trigger a solid reaction from the institutions, and that could be a welcome exception to what would be expected based on our ranking.
The list of locations where the “seismic memory” has been well cultivated certainly includes many other centers, especially in central and southern Italy: but the effectiveness of these virtuous behaviors will receive confirmation only from the forthcoming earthquakes.
The case of Norcia seems quite special. The famous building regulations enforced by the Papal State (1859) seem to have contributed since then to limit the damage, even in the case of the 1979 earthquake. Viceversa, for mysterious reasons Norcia was included in the seismic code only in 1962. In 2016 it suffered more damage outside the walls than inside. Do you have an opinion on this?
Norcia is ranked 676° place in our classification, mainly by virtue of the 1979 earthquake, but had previously suffered intensity VIII or larger effects in 1730, 1859 and 1879.
The case of Norcia is indeed quite unique. The “fortune” of Norcia towards earthquakes – if the term is granted to me, being perhaps inappropriate in view of what has happened in the city over the past few months (5) – is largely due to two symbolic earthquakes, those of 1859 and 1979, both with a magnitude of around 5.8, and to a sort of wake-up earthquake, that of 1997. Let me explain it better.
Following the 1859 earthquake the prelate Arcangelo Secchi and the architect Luigi Poletti prepared a very accurate analysis of what had happened, accompanied by recommendations on the reconstruction collected in the famous “Building Regulations” approved between the end of 1859 and the spring of 1860. It for this reason that the 1979 earthquake found a building patrimony that on average was substantially more hard-wearing than that of the surrounding towns, although the lesson imparted by the earthquake of 120 years before had perhaps already been partly lost. After 1979 Norcia was rebuilt with a great commitment, both by residents and institutions. The 1997earthquake, whose epicenter was quite far from Norcia, was the occasion for a “recall” of what had been done after 1979, as it is done with vaccines. The nursini – the people of Norcia – have the earthquake in their DNA: and I state this with full knowledge of the facts because some of my maternal cousins were born and raised there.
I believe that Secchi and Poletti’s famous Building Regulations did play a major role; an example for all, that of the Civic Tower, which miraculously survived the 30 October 2016 earthquake. Norcia demonstrates that the lesson taught by history to the local culture may compensate for any delays in the introduction and implementation of anti-seismic codes. In Norcia the local culture has not waited for modern codes but has anticipated them, also thanks to Secchi and Poletti. We should also remember that in Italy the codes have always only affected only new buildings and those that have been significantly restored; nothing is imposed to the owners of existing buildings. In my opinion this is one of the great unresolved issues, perhaps the greatest, as well as a source of misunderstandings and ill-fated expectations.
If we want, the case you mention – that of greater damage outside the walls of Norcia compared to the historic center s.s. – is a paradoxical confirmation of the role of “historical memory” in mitigating the effects of earthquakes. Here, too, an engineer should speak in my place, but I will try to venture hypotheses, some of them quite obvious.
First of all it must be said that the value of the “historical memory” of the nursini applies only to the “historical component” of the building stock. This statements seems redundant, but in fact what could be the value of “historical memory” for a condominium built in the 1980s, very different from the constructive style of the city center but rather similar to what you see in many urban suburbs of Italy?
The construction style is also the basis of my second hypothesis, stemming from the evidence that a load-bearing masonry building can also defend itself very well from earthquakes, provided that it is well built or renovated, according to the best practices in use in the various epochs. Recall that the 30 October shock was a Mw 6.5 earthquake located just below the center of Norcia: the accelerations observed were very significant, to the point of making the performance of masonry buildings truly extraordinary. In condominiums built outside the walls, instead – but I insist that this is the opinion of a geologist – once again we have seen that in normal reinforced concrete buildings the performance of the supporting structure can also be very different from that of of infills and of any accessory structures. This means that the building is unlikely to collapse, unless there are evident flaws in its design, but also that non-structural damage can be so burdensome as to make it convenient to demolish and rebuild: a paradox which I believe has been addressed in the new Norme Tecniche sulle Costruzioni 2018 (Technical Standards for Construction, or NTC18), at least for the future.
Have you had the chance to discuss this study with some seismic engineer, or have you received any reaction from that environment? How do you think your results can be used, and above all by whom?
We have received words of encouragement both from various engineers to whom we have submitted the first version of the manuscript, and from the public to whom we have presented the study in the most diverse occasions. But that’s all, because there have not been other reactions, at least for now – and I refer above all to those of the institutions. We also tried to establish a relationship with the Casa Italia Mission Structure, when it was still directed by Prof. Azzone, but also in this case there was no reaction. Evidently Casa Italia does not share with us the need to set priority criteria for risk mitigation interventions soon: interventions that the Structure is not implementing anyway, if not in the ten symbol-building sites that will be opened in as many symbol cities.
We believe that earthquake mitigation must combine excellent scientific assumptions – and in Italy we believe we have both a rich heritage of seismicity data, and an excellent expertise to make the best use of them – with much pragmatism as regards how and where invest any resources that may become available for seismic improvement. We also believe that the Sisma Bonus (“Seismic Bonus”) could be a useful tool, but only on condition that the criteria for assigning its benefits are drastically reviewed (and that the Sisma Bonus is made overall more “attractive”, by reviewing the delivery mechanisms: but on this I leave the floor to the experts of financial things). In particular, we maintain that a ranking of priorities must be drawn up between the various municipalities and the various aggregates of buildings, choosing them on the basis of their presumable vulnerability – on the basis of hypotheses such as those formulated by us – or real – on the basis of punctual findings, even if expeditious.
Launching this process harmoniously requires a solid control room, which I believe should include researchers, representatives of the professional associations involved, ISTAT officials, as well as institutional representatives of various origins (the Italian Civil Protection, the Ministry of Economic Development or MiSE, the National Association of Italian Municipalities etc.). For over two years I thought and hoped that this control room could coincide with the structures of Casa Italia, but today it is clear to me that I was wrong.
Finally – last but not least – it is necessary to launch a multi-year approach to the mitigation of seismic risk; an approach which at least on such an important issue breaks down the endemic “five years perspective” (when it’s good) that from always characterizes the governments of the Belpaese. But the ability to effectively plan the future is not one of the traditional virtues of the Italians, and therefore I fear we will not go very far on this side.
As a researcher I can clearly distinguish what makes sense from what could only be written in a book of dreams. However – and with this I close the circle you opened with the first question – I would like to dedicate the next few years to promote a change of course on how these issues are coped with in Italy today. I consider it a moral duty of my generation of seismologists; a generation stemming from the immense – and certainly avoidable – catastrophes of Friuli and Irpinia, and from the subsequent birth of a modern and effective Civil Protection.