Influence of local site effects on seismic risk maps and ranking of Italian municipalities

doi:10.21203/rs.3.rs-2061611/v1

Download PDF

Research Article

Influence of local site effects on seismic risk maps and ranking of Italian municipalities

https://doi.org/10.21203/rs.3.rs-2061611/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 03 Feb, 2023

Read the published version in Bulletin of Earthquake Engineering →

You are reading this latest preprint version

The latest studies concerning seismic risk assessment in Italy were developed without considering the site geo-lithological effects, thus limiting the assessment to rock soil type. In this study, for the evaluation of site amplification effects, we used the results of recent works based on the Italian seismic microzonation data. The site amplification factors have been assigned to the chief town of each of the 7,715 Italian municipalities, assuming that the building stock is concentrated in that small area. The PGA amplification ranges from 1 to 2.2 bringing to an average increase of 76% in the hazard. The seismic risk has been evaluated using the probabilistic hazard assessment adopted by the Italian building code and recent vulnerability/exposure models developed for Italy. The residential building stock was subdivided into 5 vulnerability classes (3 for masonry and 2 for concrete) derived from the national census data and further refined with the construction age and building height. The results show that, for a rock site condition, the loss estimates are comparable with those of previous works and with the historical data retrieved from a careful analysis of the major Italian earthquakes in terms of casualties, homeless, and economic damage. The introduction of the site effects brings to a significant increase in the expected losses, mainly due to the underestimation, in the adopted fragility curves, of the PGA taken from shakemaps not accounting correctly for the site amplification. To support the implementation of risk reduction programs, a seismic risk index is introduced, allowing a municipality ranking by risk, a cost estimation of the seismic retrofitting in the highest risk municipalities, and an assessment of the consequent risk reduction.

Seismic hazard

seismic risk

site effects

fragility curves

loss assessment

risk reduction programs

Italy is one of the countries with the highest seismic risk in Europe, due to the frequency of earthquakes that have historically affected its territory and to the intensity that some of them have reached, causing extensive social and economic losses, mainly due to the high population density and to the considerable fragility of the building stock (Valensise et al. 2017). It follows that the major issue in Italy does not concern the design of new buildings, but rather the high vulnerability of existing buildings which may be reduced with appropriate and well-known techniques as strengthening, retrofitting, and base isolation.

The Italian annual expenditure for the recovery from past seismic events (an average of 3 billion Euro) is larger than for the prevention of losses for future ones. Considering the very large amount of funds required to strengthen/retrofit the whole Italian building stock (estimated in the order of trillion Euro), it is essential to establish a priority scale of intervention through a well-grounded and reliable seismic risk assessment providing the extent of damage to people and buildings expected for future earthquakes. These are also key factors for insurance and reinsurance companies (Goda et al. 2014).

The first seismic risk maps developed in Italy at national scale date back to 1996 (GNDT-ING-SSN, 1996, Dolce et al., 1996). These maps represented the first attempt to express the seismic risk as a combination of hazard, vulnerability and exposure, even if they had not been included in peer-reviewed journal articles. The exposure model represented the main novelty of that study, on the basis of population and buildings data collected in the 1991 census by ISTAT (Italian National Institute of Statistics). The latest census datasets (ISTAT 2001, 2011) do not differ significantly from the study of 1991 and are still used in recent risk assessments for their homogeneity and completeness over the national territory.

Crowley et al. (2009) provided a complete report of the studies on Italian seismic risk maps up to 2009. The report compares several works, such as the risk maps of the Italian Seismic Service (Lucantoni et al. 2001) making use of seismic hazard assessments (Albarello et al. 2000), damage probability matrices and fragility curves (Sabetta et al. 1998, Di Pasquale et al. 2005); the SAVE project (Zuccaro 2004); the risk assessment performed with new analytical vulnerability models (Borzi et al. 2008). More recently, several other studies were published, such as those of Rota et al. (2011), Asprone et al. (2013), Zanini et al. (2019), and the study realized by the Global Earthquake Model (GEM) foundation (GEM 2019; Silva et al. 2020).

The latest comprehensive study concerning Seismic risk studies of Italian residential buildings is provided by Dolce et al. (2020). In that work six research units, coordinated by the Italian Department of Civil Protection, collaborated to update the risk maps of the Italian residential building stock providing six different vulnerability/exposure models (four for masonry buildings and two for RC buildings). The paper describes in detail the adopted methodology and presents the main results in terms of expected damage and impact measures (unusable buildings, homeless, casualties, direct economic losses). As acknowledged by the authors themselves, the work of Dolce and co-authors still presents some drawbacks that could underestimate the predicted losses, in particular regarding the consideration of the uncertainties and the reference to a soil of type A only (outcropping rock characterized by flat topography).

The aim of this study is to evaluate the effects on the Italian seismic risk maps of the inclusion of the site geo-lithological amplification in the hazard assessment (Falcone at al. 2021; Mendicelli et al. 2022) and of the epistemic uncertainty in the hazard. Furthermore, a seismic risk index is introduced that allows a ranking by risk of the Italian municipalities, a cost estimation of the seismic retrofitting of the most vulnerable buildings, and an evaluation of the consequent risk reduction.

Seismic risk can be defined as a measure of the damage expected at a given region of interest in a certain time interval and can be represented by the convolution of hazard, vulnerability, and exposure (Fig. 1). Seismic hazard expresses the probability of occurrence of a given earthquake or ground motion and can be obtained by applying the Probabilistic Seismic Hazard Assessment (PSHA, Cornell 1968) which provides seismic hazard curves giving λ_IM (frequency of exceedance) as a function of im, (intensity measure of the seismic events, in our case PGA). An example of hazard curves is illustrated in the left plot of Fig. 1 for the Italian municipalities of Firenze, L’Aquila and Napoli. The seismic vulnerability (in our case limited to residential buildings) represents the potential of a structure to undergo a given level of damage and it is generally described through fragility curves, expressing the probability of reaching different levels of damage given the PGA (central plot in Fig. 1). The exposure refers to all the assets which are subjected to the seismic risk in a particular area. In case of the building stock, a building inventory, providing the number of buildings, dwellings, residents and the percentage distribution in the different vulnerability classes, is required (right plot in Fig. 1).

The mean annual rate λ_k of reaching a damage state D_j may be expressed as:

$${\lambda }_{k}=\underset{0}{\overset{\infty }{\int }}P\left({D}_{k}\right|im)\bullet |d{\lambda }_{IM}\left(im\right)|$$

which numerically becomes:

$${\lambda }_{k}=\sum _{i=1}^{n-1}0.5\bullet \left[P\left({D}_{k}|{im}_{i})+P({D}_{k}|{im}_{i+1}\right)\right]\bullet \left[{\lambda }_{IM}\left({im}_{i})-{\lambda }_{IM}\left({im}_{i+1}\right)\right)\right]$$

where P(D_k|im) is provided by the fragility curve. It expresses the probability of reaching a damage state ≥ D_k as a consequence of a seismic event with seismic intensity im for buildings of a particular class of vulnerability. λ_IM is the mean annual exceedance frequency of the intensity im.

The calculation should be repeated for each vulnerability class of the building inventory and the results should be combined considering the proportion of each class in the considered asset and summed over the different classes to get the probability of reaching a given damage state in a given site or municipality. The λ_k can be used to compute the probability p_k to reach the damage state D_k in a time interval of t years under the hypothesis that the occurrence of events is a Poisson process:

$${p}_{k}\left(in t years\right)=1-{e}^{-{\lambda }_{k}\bullet t}$$

Equation (3) provides the unconditional seismic risk in a time interval t (expressed in years) for a damage state D_k. For small values of λ_k, as is usually the case, and t = 1 year, p_k ≈ λ_k. The mean annual probability of reaching the different damage states for all the buildings classes can be used to provide an estimate of the consequences of an earthquake in terms of economic, human, and social losses.

In our study, as hazard curves referred to outcropping rock, we used the probabilistic assessment MPS04 (Stucchi et al. 2004, 2011) that is the current reference for seismic hazard studies in Italy and is adopted by the national building code (NTC, 2018); for vulnerability we adopted the fragility curves implemented by Rosti et al. (2020a, 2020b); for exposure we used the building and population census dataset (ISTAT 2011) giving information on construction age, building material, and number of storeys of the buildings.

A substantial improvement in the evaluation of the geo-lithological site amplification effects in Italy has been given by the recent works of Mori et al. (2020) and Falcone et al. (2021). The first work, following the methodology suggested by Iwahashi et al. (2018) for a site classification using digital elevation models (DEM), has derived a V_S30 map for Italy starting from the geomorphological classes and integrating a large amount of data from the Italian seismic microzonation dataset (DPC, 2018; www.webms.it), consisting of 16,000 profiles of shear wave velocity from geophysical surveys and 44,000 logs from continuous coring boreholes. Mori and co-authors found a proxy dependence of V_S30 on slope gradient and elevation through a regression model for each geomorphological class. The estimation of V_S30 for the Italian territory provided by their study is particularly accurate and provides a good correlation between V_S30, morphological classes and subsoil lithologies.

In Fig. 2 the new V_S30 map from the study of Mori and co-workers is compared with that of Michelini et al. (2008) used so far for the implementation of the shakemaps which have been widely used by the Italian engineering community for the evaluation of fragility curves in building vulnerability assessment. It is worth noting that Sardinia is not included in the present work because of its negligible value of seismic hazard (Stucchi et al., 2004). The methodology of Michelini and co-workers is based on soil classification derived from the 1:100,000 geological map of Italy (Amanti et al. 2008) and from the topographic slope (Wald and Allen, 2007). The differences between the two maps are substantial. In particular, in the map of Michelini et al., most of the territory is classified as having a V_S30 greater or equal to 800 m/s whereas in the new map less than 3% of the territory has a V_S30 greater than 760 m/s and about 50% is in the range 360–480 m/s.

In later studies (Falcone et al., 2021; Mendicelli et al. 2022) the geo-lithological amplification factors are based on site-specific data using a one-dimensional numerical approach for the seismic site response study. Detailed site data were retrieved from the Italian database for seismic microzonation (DPC, 2018) and were grouped according to the geomorphological clusters described in Mori et al. (2020). The characterization of homogeneous areas for lithological and geophysical properties was carried out through site investigations based on Vs profiles with depth and lithology distributions. The authors created a code, based on the equivalent visco-elastic method in the frequency domain, which allowed to perform around 30 million one-dimensional local site response analyses.

Figure 3a shows the distribution of the AFs of the PGA, for a 475 years return period, assigned to the chief town of each of the 7,715 Italian municipalities excluding Sardinia (ISTAT, 2011). The PGA value at the municipality is derived, as for the national building code (NTC, 2018), with an interpolation of the four grid points of the MPS04 map (Stucchi et al., 2004) closest to the coordinates of the chief town. The basic assumption is that the building stock is concentrated in the relatively small area of the chief town. Only 16 municipalities (i.e. 0.19% of the total) have an AF slightly lower than 1 corresponding to a rock soil type. The PGA amplification factor ranges therefore from 1 to 2.2, with the highest values in Northwestern Italy and the lowest in the eastern parts of the Po Plain and Puglia region. Figure 3b shows the AFs calculated (Falcone et al., 2021) in the range 0.4–0.8 s of the elastic response spectrum. Moving to higher natural periods, the territorial distribution of the AFs changes significatively with the highest values in the Po Plain. In fact, according to the simplified rule f₀ = V_S/(4·H), the higher the thickness H of a deposit, the lower the fundamental frequency f₀. Therefore, the highest amplification for the deep deposit of the Po Plain is reached for the period range 0.4–0.8 s rather than for PGA.

In this study the AFs have been calculated for each of the 9 earthquake return periods (Tr) given in Stucchi et al. (2011) and in the national building code (NTC, 2018). Figure 4 shows different examples of the AF versus V_S30 trend. In Fig. 4a, the PGA-AFs of the Italian municipalities are plotted for the hazard levels corresponding to three return periods. The values display a linear increase up to about 300–400 m/s becoming scattered and decreasing at lower V_S30 values due to the nonlinear soil behavior. This effect is more pronounced for a Tr of 2475 years due to its greater PGA as shown by the interpolating lines. In case of a spectral period interval 0.4–0.8 s. (Fig. 4c), the AFs are less dispersed and increase up to a value of about 5 at 150 m/s. Figures 4b and 4d show the AF versus V_S30 trend foreseen by the Italian building code (NTC 2018) in case of two cities, Milano and L’Aquila, with low and high seismic hazard, respectively. The four soil classes considered by the code have been assigned to the following V_S30 reference values: (A) 900 m/s; (B) 600 m/s; (C) 300 m/s; (D) 150 m/s. The AFs of the Italian building code, different for each municipality, are comparable with those of Falcone et al. (2021) both for the values of PGA and for those relating to the interval 0.4–0.8 s (0.6 s in Fig. 4d). In Figs. 4b and 4d are also reported the AFs proposed by Michelini and co-workers, used for the implementation of the shakemaps and the fragility curves, as discussed previously. It is evident that, in particular for high accelerations, the AFs are strongly underestimated. This will result in an overestimation of the damage probability in the fragility curves as described in the following sections.

Figure 5 shows the effect of the introduction of the site amplification in the hazard map MPS04 (Stucchi et al., 2004), with PGA values assigned to the chief town of each of 7,715 municipalities (Sabetta, 2021). The maximum PGA increases from 0.28 g (municipality of Ferla in east Sicily) to 0.47 g (municipality of Scigliano in Calabria). The average increase due to the geo-lithological amplification over the national territory is 76%.

The estimation of the risk requires to use the hazard curves with the relative epistemic uncertainty (Meletti and Montaldo, 2007) as shown in Fig. 6a for the cities of Biella and L’Aquila. The nine points for which the curves are provided have been interpolated, at a step of 0.001 g, with a best fit function as shown by the correlation index R² reported in Fig. 6a. To avoid an overestimation of the losses due to very low levels of ground shaking associated with high probability of exceedance and non-zero damage probability, the hazard curves were truncated at values lower than 0.03 g (red line in Fig. 6a). The uncertainty in the mean hazard, represented by the ± σ dashed lines, is in the order of 10–15% for high hazard municipalities and 20–25% for low hazard ones. It results rather low, probably due to the sixteen branches logic tree used in the implementation of the MPS04 map (Stucchi et al., 2004) which does not include enough variability. To include the site effects, the hazard values for rock have been multiplied by the AF varying with Tr, as shown in Fig. 6b for some representative cities. For the sake of convenience, the uncertainty related to the geo-lithological site amplification was not accounted for, hence, only the 50th percentile of the AF distribution was retrieved from the study of Mendicelli et al. (2022).

According to the 15th ISTAT Census of population and housing (ISTAT, 2011), in 2011 there were 12,187,698 residential buildings for a total of 31,208,161 dwellings. With respect to the age of the housing stock, it emerges that 53.7% of dwellings have been built before 1970, a further 31% was built in 1971–1990 and 7.4% in the period 1991–2000. The remaining 7.9% of the housing stock was built between 2001 and 2011. The homes occupied by resident people are 24.1 million (23.4 million without Sardinia), equal to 77.3% of the total, while seven million are unoccupied (holiday homes, homes occupied by non-residents, empty homes). For evaluating the exposure, we considered the census data concerning the construction age, number of storeys, and building material (masonry M, reinforced concrete RC, other O). As the adopted fragility curves (Rosti et al., 2020a, 2020b) are relative only to M and RC, it was necessary to subdivide the category that ISTAT indicates as O, i.e., all the buildings with a mixed load-bearing structure and a small percentage of steel and timber buildings, between the M and RC classes. To do this, it has been used the criterion suggested by Dolce et al. (2020) based on the percentage of O in each town, the period of construction and the number of floors. Using the classification criteria reported in Rosti et al. (2020a, 2020b), the residential buildings have been divided into twelve vulnerability classes, three for masonry (A, B, C1) further subdivided in two height classes, and two for reinforced concrete (C2, D) further subdivided in three height classes. Regarding the masonry, the breakdown into classes is based on the building age and number of floors, as illustrated in Table 1, where L and MH are the classes of height considered (1–2 floors and 3 or more floors respectively).

Table 1

Distribution of the masonry building stock into the vulnerability classes as a function of building age and class of height (Rosti et al. 2020a).
Class of height	Vulnerability class	Construction period
Class of height	Vulnerability class	< 1919 (%)	1919-45 (%)	1946-61 (%)	1962-71 (%)	1972-81 (%)	> 1981 (%)
L	A	86	45	9	5	0	0
	B	0	44	59	4	0	0
	C1	14	11	32	91	100	100
MH	A	97	22	0	0	0	0
	B	0	78	75	18	0	0
	C1	3	0	25	82	100	100

The RC, according to Rosti et al. (2020b), has been divided in two classes assuming the year 1981 as a turning point in the application of the Italian seismic legislation. The vulnerability class C2 includes buildings designed essentially for gravitational loads (≤ 1981) while the vulnerability class D includes buildings with seismic design (> 1981) as illustrated in Table 2.

Table 2

Distribution of the reinforced concrete building stock into the vulnerability classes as a function of class of height and building age (Rosti et al. 2020b)
Class of height	Vulnerability Class
Class of height	C2 ≤ 1981 (%)	D > 1981 (%)
L	55	61
M	26	27
H	19	12

Each vulnerability class has been further refined based on the height of the building (L = 1–2 floors, M = 3 floors, H = 4 or more floors). Classes A, B, C1 correspond to masonry of poor, average, and good quality respectively. Classes C2, D correspond to RC building without and with seismic design respectively. For the risk assessment, given that the population exposure is relevant, the same class breakdown has been applied to the dwellings belonging to each building, because the ISTAT census database distinguishes between inhabited and non-inhabited buildings in terms of dwellings. Since the number of dwellings in a building is highly variable, with typically more dwellings in a RC building with respect to a M building, the classification in terms of dwellings is more appropriate to provide a clearer picture of the actual vulnerability distribution and it has been used in the loss assessment. Figure 7 shows the distribution of the residential dwellings in the vulnerability classes regardless of the building height. The masonry dwellings represent 46% of the total and the largest vulnerability class (32%) is C2 (RC built before 1981).

Figure 8 shows the percentage distribution of the dwellings belonging to the highest and lowest vulnerability classes regardless of the building height, respectively class A (masonry built mostly before 1945) and class D (RC built after 1981). The most vulnerable masonry houses are concentrated in North-Western, Central, and Southern Italy (Fig. 8a). In Fig. 8b, it can be seen (large number of seismically designed buildings) the effect of the reconstruction after the Irpinia 1980 earthquake (south-east of Naples) and the Belice 1968 earthquake (south of Palermo) which after 1981 was still in progress. The census data refer to 2011, so the beneficial effects given by the reconstruction after the 2009 L’Aquila earthquake are limited.

Fragility curves represent a fundamental tool for estimating the possible damage resulting from seismic events. With reference to a certain type of building, they express the probability of overcoming the different Damage States (DS) as a function of the intensity of the earthquake shaking. In our study we adopted the fragility curves proposed by Rosti et al. (2020a, 2020b). These two studies propose an empirical model of fragility obtained by statistically processing the damage data of Irpinia 1980 and L'Aquila 2009 earthquakes which were selected for the availability of complete post-earthquake field surveys (Dolce et al., 2019) including undamaged buildings, and for the lack of significant aftershocks avoiding the problem of the cumulated damage. Furthermore, the cited works adopt empirical models that are more suitable for masonry buildings representing almost half of the building stock in Italy.

The curves are described by a cumulative lognormal distribution and DS are taken from the five damage states (from DS1 slight or negligible damage to DS5 collapse) of the European Macroseismic Scale (EMS-98, Grünthal 1998). The main shortcoming is that these fragility curves are still based on the shakemaps of Michelini et al. (2008). The comparison of the site AFs used in this work (Fig. 4a) with those of Michelini 2008 (Fig. 4b), reveals that the latter are strongly underestimated leading to an overestimation of the damage probability. Figure 9 compares the fragility curves for some DS and vulnerability classes previously described. It is worth noting that in some cases the curves intersect, e.g., the curves C2-H (concrete without seismic design, 4 or more floors) at high PGAs and for damage states 3 and 5, exceed the values of poor-quality masonry A-MH.

The risk calculation has been performed for each of the 7,715 Italian towns according to Equations (2) and (3) and considering the dwellings distribution in the different vulnerability classes described in section 4. Table 3 shows the resulting estimates in terms of the number of dwellings (thousands) affected by each damage state in a time frame of 1 and 50 years. The results, if the site amplification is not considered, are quite comparable with those provided in the study of Dolce et al. (2020). Figure 10 shows the unconditional risk in a period of one year corresponding to the average expected probability of damage states DS1 and DS5 including the site amplification. The risk of collapse (DS5) has lower values with respect to DS1 and it is concentrated in Central and Southern Italy where both the amplified PGA and the building vulnerability are the highest.

Table 3

Expected number of dwellings affected in 1 and 50 years by the considered five damage states compared with the results of Dolce et al. (2020), thousands of dwellings.
	DS1	DS2	DS3	DS4	DS5
This study 1 year without site amplification	103.6	40.0	20.3	6.3	1.6
Dolce et al. (2020) 1 year	143.1	38.7	17.8	6.1	2.1
This study 50 years without site amplification	4412.5	1853.4	974.2	312.4	77.5
Dolce et al. (2020) 50 years	4199.7	1436.0	783.0	290.9	103.6

The definition of specific impact indicators based on seismic risk assessment can be very useful in the field of civil protection, with possible application to response and mitigation strategies, as well as for emergency planning.. In this study, we estimated the number of dwellings resulting unusable on the basis of existing risk studies for Italy (Lucantoni et al., 2001; Bramerini and Di Pasquale, 2008; Zuccaro and Cacace, 2011, Dolce et al. 2020). In particular, for the unusable dwellings, we adopted the formulation given by Dolce et al. (2020) for the long-term unavailability which gives values that are lower and closer to the data from field surveys. The number of homeless was derived from the resident people in the unusable dwellings also considering the people not counted as victims (99% DS4, 90% DS5). The probability of injury or death of the building occupants has been calculated as a function of dwelling damage states DS4 and DS5, as proposed by Lucantoni et al. (2001) and subsequently developed by Zuccaro and Cacace (2011) and Dolce et al. (2020).

The following impact indicators have been defined for each Italian municipality considering the number of dwellings and resident people belonging to each damage state:

expected number of unusable dwellings in the long term = 0.6 ∙ DS3 + DS4
expected number of homeless = 0.6 ∙ DS3 + 0.99 ∙ DS4 + 0.9 ∙ DS5
expected number of victims = 0.01 ∙ DS4 + 0.1 ∙ DS5
expected number of injured = 0.05 ∙ DS4 + 0.3 ∙ DS5

Figure 11 shows the percentage of dwellings of each municipality expected to become unusable in a period of 1 year. If the site amplification is not included, the maximum values are 0.4–0.7%; when site effects are considered, the percentage of unusable dwellings, in some towns of Central Italy, reaches values higher than 1% per year.

Figure 12 displays the expected number of victims in each municipality in a period of 1 and 50 years (Eq. 3), without considering the site amplification. Being absolute and not percentage values as in the previous maps, the effect of the population exposure becomes prevalent and the higher values correspond to the big cities as Naples, Rome, Palermo, Catania, Messina. Table 4 reports the expected losses over the national territory in a time frame of 1 year obtained averaging (probability of damage state) or summing (losses) the values relative to the single municipalities.

Table 4

Unconditional risk in a time frame of 1 year: average expected values over the Italian territory
	annual prob. DS1	annual prob. DS5	unusable dwellings (long-term)	homeless (60%DS3 + 99%DS4 + 90%DS5)	victims (1%DS4 + 10%DS5)	injured (5%DS4 + 30%DS5)
Dolce et al. 2020			15,635	78,602	505	1,744
This study without site effects	3.80E-03	8.51E-05	18,534	48,921	540	1,934
This study with site effects	8.23E-03	3.47E-04	65,683	176,910	2,493	8,686
Relative increase due to amplification	116%	307%	254%	262%	361%	349%

If the site amplification is not included, the estimated losses are comparable with those given in Dolce et al. (2020) except for the number of homeless people, particularly high in that work because they were calculated considering the inhabitants of the unusable dwellings expected both in the long term (0.6∙DS3 + DS4) and short term (0.4∙DS2 + 0.4∙DS3). It is worth noting that, due to the shape of the fragility curves, the increase in damage estimation, going from 116% for DS1 to 307% for DS5 if the site amplification is included, is larger than the average increase in PGA (76%). When the site effects are included, the loss assessments result largely overestimated both with respect to previous studies and to real values surveyed in the field (see following section). As discussed previously, this is due to the adopted fragility curves which use a seismic input (shakemaps) where the site geo-lithological amplification is underestimated and underlines the need of a re-evaluation of fragility curves with the new site amplification factors provided by Falcone et al. (2021) and Mendicelli et al. (2022).

A crucial element for the assessment of the reliability of the loss estimates is the comparison with the real values surveyed in Italian earthquakes. Table 5 shows information about medium-strong earthquakes of the last 54 years, which is the only period for which reliable data about homeless and economic damage are available. The last column of the table reports the total cost in Million €, discounted to 2014, incurred by the Italian government for the emergency management and the reconstruction of structures, infrastructures and cultural heritage.

Table 5

Victims, homeless and economic cost relative to 16 earthquakes of the last 54 years with an intensity ≥ VII degree of MCS scale.

Source: *catalog CFTI5Med Guidoboni et al. (2019); **Web; ***specific study; ****Italian civil protection. The last column reports the total cost, in Million € discounted to 2014, incurred by the Italian government for the reconstruction and the emergency management. Data in italic are estimated from a linear regression with the number of homeless
Earthquake	Date	Mw	Io MCS		Victims	Homeless	Source	Cost M€
Belice	15/01/1968	6.5	X		290	100,000	***	9,179
Tuscania	06/02/1971	5.2	VI-VII		20	4,000	*	1,807
Parmense	15/07/1971	5.7	VIII		1	1,000	*	704
Ancona	14/06/1972	5.5	VIII		0	30,000	*	7,112
Friuli	06/05/1976	6.5	X		978	200,000	*	18,540
Valnerina	19/09/1979	5.8	VIII-IX		5	25,000	*	6,283
Irpinia	23/11/1980	6.8	X		2,914	280,000	*	52,026
Appennino abruzzese	07/05/1984	5.8	VIII		0	6,000	*	2,381
Sicilia sud-orientale	13/12/1990	5.4	VII		12	15,000	*	4,439
Umbria-Marche	26/09/1997	5.7	VIII		11	38,000	***	13,463
Appennino Lucano	09/09/1998	5.5	VII		1	650	**	525
Molise	31/10/2002	5.7	VII-VIII		30	3,000	**	1,400
Salò	24/11/2004	5.0	VII-VIII		0	2,000	**	1,128
L'Aquila	06/04/2009	6.3	IX-X		309	67,460	****	17,458
Emilia	29/05/2012	6.0	VII-VIII		17	16,520	**	8,171
Italia centrale	24/08/2016	6.2	X		299	31,760	****	13,163
				TOTAL	4,887	820,390		157,778
				per year	92	15,479		2,977

Cost data up to 2002 are taken from a report of the Italian House of Representatives (Camera dei Deputati, 2009). Data in italic are estimated from a linear regression with the number of homeless for the 8 earthquakes for which the cost is available. Data relative to the last three earthquakes, since 2009, come from a report of the Italian Senate (Senato della Repubblica, 2017). A more in-depth analysis of these data (not final because the reconstructions are still in progress) reveals that 85% of the resources were destined to the reconstruction of damaged or collapsed buildings, 8% to productive activities, 4% to public administrations and 3% to emergency management.

The count of the victims, which should be more reliably reported in the historical chronicles with respect to homeless and economical losses, is highly variable depending on the period considered. The number of victims per year, which is 92 for the last 54 years, increases up to 994 if counting since the beginning of XX century because in that case are included Messina 1908 and Avezzano 1915 earthquakes summing up 113,000 victims, about 94% of the total. Table 6 shows the victims of destructive earthquakes (intensity ≥ X degree of MCS scale) considering the entire catalogue (Guidoboni et al., 2019) since 1169. It is interesting to note that, as shown in Fig. 13a, these destructive earthquakes are concentrated essentially in Southern Italy. Considering that the Italian seismicity does not seem to have had large modifications in the last 1000 years, as illustrated in Fig. 13b, the variation is mostly due to the social, demographic, and building characteristics of the historical period considered and to the site where the earthquake happened (country or city). The most reliable value is probably that reported in Table 6, averaging the short period variation, and giving 407 victims/year.

Table 6

Victims relative to 35 destructive earthquakes of the last 853 years with an intensity ≥ X degree of MCS scale.

Source: *catalogue CFTI5Med Guidoboni et al. (2019); **Web; ***specific study; ****Italian civil protection.
Year	Epicentral area	Io	Mw	Victims	Source
1169	Sicilia orientale	X	6.4	15,000	*
1328	Norcia	X	6.4	2,000	*
1349	Lazio meridionale-Molise	X	6.6	2,500	*
1456	Molise	XI	7.2	25,000	*
1461	Aquilano	X	6.4	150	**
1542	Siracusano	X	6.8	200	*
1561	Vallo di Diano	X	6.8	584	***
1627	Gargano	X	6.7	5,000	*
1638	Crotonese/Calabria	X	6.9	30,000	*
1659	Calabria centrale	X	6.6	2,035	*
1688	Sannio	XI	7.0	10,000	*
1693	Sicilia orientale	XI	7.4	54,000	*
1694	Irpinia-Basilicata	X	6.8	4,820	*
1695	Asolano	X	6.5	150	*
1702	Beneventano-Irpinia	X	6.5	400	*
1703	Aquilano	XI	6.7	9,670	*
1706	Maiella	X-XI	6.8	2,400	*
1732	Irpinia	X-XI	6.6	1,940	*
1781	Appennino marchigiano	X	6.4	300	*
1783	Calabria	XI	6.6	35,170	*
1805	Molise	X	6.6	5,573	*
1832	Crotonese	X	6.6	234	*
1851	Basilicata	X	6.4	1,000	*
1854	Cosentino	X	6.2	500	*
1857	Basilicata	XI	7.0	19,000	*
1908	Calabria-Messina	XI	7.1	80,000	*
1915	Avezzano	XI	7.0	33,000	*
1919	Mugello	X	6.3	100	*
1920	Garfagnana	X	6.5	171	*
1930	Irpinia	X	6.6	1,404	*
1968	Valle del Belice	X	6.3	290	***
1976	Friuli	X	6.5	978	*
1980	Irpinia-Basilicata	X	6.9	2,735	*
2009	L'Aquila	X	6.3	309	****
2016	Amatrice	X	6.2	299	****
		TOTAL		346,912
		Per year		407

The total direct economic losses may be computed based on loss parameters related to damage repair and to the building surface reported in the inventory data. The cost parameters are calibrated as the percentage cost of repair or replacement for each structural damage state and equal to 0.02, 0.1, 0.3, 0.6, 1, for DS1, DS2, DS3, DS4, and DS5, respectively, with a unit cost of 1,350 €/m² including technical expenses and VAT (Di Ludovico et al., 2017; Dolce et al., 2020). If the site amplification is not considered, the estimated cost/year for repair/reconstruction of damaged buildings, is 2.43 billion €, similar to that provided in the work of Dolce and co-workers summing up to 2.13 billion €/year. If site effects are considered the cost increases up to 7.3 billion €/year, resulting overestimated but to a lesser extent compared to victims and homeless. Figure 14 shows the mean annual expected economic losses (Million €) over the national territory with and without considering the site amplification effect. Being absolute values as in Fig. 12, the greater losses are concentrated in the largest cities even if characterized by relatively low seismic hazard. To this end, Table 7. shows, for three Italian cities, the effect of the exposure in terms of inhabitants and dwellings on the absolute values of economic loss and victims compared with hazard (probability of exceeding a PGA of 0.2 g) and risk (risk index as defined in section 10). Rome and Naples, even if affected by a rather low hazard and risk, have the highest values of economic loss and victims, both with and without site amplification, because of their large exposure values. Campobasso, a medium-sized city northeast of Naples, has a definitely higher hazard and risk (columns 2 and 5 in Table 7) but a much lower number of victims and damages expected per year.

Table 7

Effect of the exposure in terms of inhabitants and dwellings on the absolute values of economic loss and victims compared with hazard (probability of exceeding a PGA of 0.2 g) and risk (risk index as defined in paragraph 10) for three Italian cities, with and without site amplification (soil, rock).
City	probability exc. 0.2 g	Inhabitans	Dwellings	Ranking by risk index (see par.10)	loss/year million €	victims/year
Roma rock	8.00E-04	2,563,863	1,135,983	2181	153.5	16
Roma soil	3.70E-03			1063	645.4	133
Napoli rock	1.50E-03	955,158	347,470	661	83.8	23
Napoli soil	5.50E-03			449	244.7	102
Campobasso rock	3.20E-03	48,556	18,981	238	8.2	2
Campobasso soil	1.25E-02			70	24.6	10

Table 8 presents a comparison of the estimated losses obtained in this study with those derived from historical surveyed data shown in Tables 5 and 6. The loss estimates have been calculated also considering the epistemic uncertainties in the hazard curves: 50°, 16° and 84° percentiles (Stucchi et al., 2011, Meletti and Montaldo 2007). The values including site amplification are overestimated due to the reasons discussed previously, whereas those without amplification are comparable, as order of magnitude, with the historical data. In fact, considering the large uncertainties in the data used for risk assessments and even in the historical data, the estimate of the order of magnitude of the expected losses is the real goal of any risk evaluation (in our case hundreds of victims, tens thousands of homeless and some billion € per year). The loss estimates referring to rock site condition are also in line with those of previous studies, both in absolute values and in territorial distribution (Lucantoni et al., 2001; Rota et al., 2011; Zanini et al., 2019; Dolce et al., 2020).

Table 8

Comparison of the loss values obtained in this study with those derived from historical surveyed data, including the uncertainty in the seismic hazard
	with site effects			without site effects
	16° perc.	84° perc.	50° perc. hazard	16° perc.	84° perc.	50° perc. hazard	historical data
victims/year	1,766	3,634	2,493	373	847	540	407
homeless/year	128,061	240,454	176,910	34,536	70,482	48,921	15,192
loss/year billion €	5.75	10.66	7.87	1.67	3.35	2.34	2.97

Although the absolute values of losses are affected by considerable uncertainties, the relative ranking of different sites and municipalities is still valid and it is particularly important to have a clear perception of the seismic risk at the national scale and to rationally implement risk reduction programs (Zanini et al., 2019). To this end, in this study we introduce a risk index calculated, for each municipality, by weighting by 2/3 the so-called life-risk (i.e., the annual percentage of homes subject to collapse) and by 1/3 the damage-risk (i.e. the annual percentage of homes subject to damage):

$$Risk index=0.33·\frac{PD}{{PD}_{max}}+0.67·\frac{PC}{{PC}_{max}}$$

where PD is given, for each municipality, by the sum of the annual probability of damage states DS2 and DS3 and PD_max is the maximum value among the 7,715 municipalities considered. PC and PC_max have been calculated in the same way considering the probability of damage states DS4 and DS5. In this way the risk index is normalized to 1. Figure 15a displays the distribution over the national territory of the seismic risk index including the site amplification. If compared to the hazard map presented in Fig. 5b, it shows that the risk is significantly modified with respect to the hazard due to the uneven distribution of the vulnerability of the dwellings (Fig. 8). Figure 15b shows the distribution of the first 100, 200 and 1,000 municipalities sorted by risk index that are concentrated in northeastern Alps, in Central Italy, in Calabria and in Eastern Sicily.

Table 9

First 10 Italian municipalities sorted by seismic risk index including and not including site amplification effect
Province	Municipality	Population	Ranking by risk index (soil)	Ranking by risk index (rock)	Ranking by vulnerability (class A)	Ranking by hazard (PGA 475 yrs. soil)
L'Aquila	Anversa degli Abruzzi	367	1	2	27	240
L'Aquila	Gagliano Aterno	255	2	3	98	207
L'Aquila	Rocca Pia	167	3	4	40	204
L'Aquila	Calascio	136	4	1	7	786
Udine	Savogna	482	5	5	47	703
Isernia	Carpinone	1,161	6	10	46	307
Udine	Pulfero	1,025	7	6	76	782
L'Aquila	Caporciano	223	8	7	64	244
Pordenone	Frisanco	633	9	12	181	113
Rieti	Micigliano	127	10	14	82	630

Table 9 shows the first ten municipalities sorted by seismic risk index and compared with the corresponding ranking in terms of vulnerability and hazard. They are, as for the majority of the first 100 municipalities except for Reggio di Calabria, Cosenza, Sulmona, and Campobasso (Table 10), small villages with a population lower than 1,000 inhabitants. Relevant cities such as Bologna, Napoli, Messina, Catania are instead included in the first thousand municipalities sorted by risk index. The effect of the site amplification seems not relevant for the ranking (columns 4 and 5) and the influence of vulnerability (represented in column 6 by the ranking in terms of percentage of class A) is higher than that of hazard (column 7, ranking in terms of PGA). Seven out of the ten highest risk municipalities are in L’Aquila province, remembering that the 2011 ISTAT census did not take into account the reconstruction after the 2009 L’Aquila earthquake that is still in progress. Table 10 shows the largest cities, in terms of population, among the first 200 municipalities with the highest seismic risk index.

Table 10

First 8 Italian municipalities sorted by population among the 200 with the highest seismic risk index
Province	Municipality	Population	Ranking by risk index (soil)	PGA 475 yrs (soil)
Catania	Catania	291,804	152	0.370
Messina	Messina	238,927	113	0.406
Reggio di Calabria	Reggio di Calabria	179,748	88	0.458
Cosenza	Cosenza	68,962	16	0.453
Bologna	Imola	67,201	195	0.379
Campobasso	Campobasso	48,556	70	0.431
L'Aquila	Sulmona	23,986	15	0.420
Catania	Sant'Agata li Battiati	9,593	186	0.375

In Table 11 are illustrated the costs of a hypothetical seismic retrofitting applied to the first 100, 200 and 1,000 municipalities sorted by risk index. The costs have been estimated considering an amount of 600 €/m² applied only to buildings of the three highest vulnerability classes, namely A and B for masonry and C2 for reinforced concrete, representing about 70% of the total dwelling surface. As a result, to seismically retrofit the first 100 Italian municipalities at risk, about 4 billion € would be needed, 20 billion € for the first 200, and about 140 billion € for the first 1,000 municipalities. It is interesting to calculate the risk reduction that can be obtained adopting this kind of intervention. For example, if the retrofit cost of 20 billion € (corresponding to about 7 years of the global expenditure normally granted by the Government for the reconstruction) was applied to upgrade the 479,614 dwellings of the first 200 municipalities at risk by one vulnerability class (A to B, B to C1, etc.) the achieved risk reduction, in terms of expected values of homeless, victims and economic loss, would be of about 57%.

Table 11

Data relative to the first 100, 200 and 1,000 municipalities sorted by risk index. The costs of seismic retrofitting are estimated considering an amount of 600 €/m² applied to buildings of the three highest vulnerability classes (A, B, C2 - about 70% of the total)
	Population	Buildings	Dwellings	Surface (m²) class A, B, C2	Retrofitting cost billion €
first 100 municipalities	404,601	103,392	164,335	1.21E + 07	7.3
first 200 municipalities	1,172,567	261,879	479,614	3.38E + 07	20.3
first 1000 municipalities	8,696,050	1,628,788	3,549,318	2.33E + 08	139.8

In this study we propose an update of the Italian seismic risk assessment considering the geo-lithological site amplification effect derived from numerical simulations based on the Italian seismic microzonation data. The amplification factors were assigned to the chief town of each of the 7,715 Italian municipalities, assuming that the building stock is concentrated in that small area. The PGA amplification ranges from 1 to 2.2 and shows an irregular pattern on the national territory. To validate the achieved results, a careful analysis of the historical data relative to the major Italian earthquakes in terms of casualties, homeless, and economic damage has been performed. If a rock site condition is considered for the entire territory, the loss estimates are comparable with those of previous works and with the historical data, both in absolute values and in territorial distribution. The introduction of the site amplification effects gives an average increase of 76% in PGA and of 100 - 350% in the probability of different levels of damage and losses. The large increase in the expected losses is mainly due to the underestimation, in the adopted fragility curves, of the PGA taken from outdated shakemaps not accounting correctly for the site amplification. A seismic risk index has been defined allowing a ranking of the Italian municipalities useful for the implementation of prevention strategies and for the risk communication to different stakeholders both in the private and public sector, such as governments, association of citizens, insurance and reinsurance companies. The first thousand municipalities at risk are mainly concentrated in Central and Southern Italy and in Northeastern Alps, reflecting roughly the hazard distribution, with significant differences because of buildings' vulnerability and exposure. Even if the absolute loss values may only be considered in terms of order of magnitude, due to the large uncertainties in hazard vulnerability and exposure, the relative ranking of different sites and municipalities is essential to implement seismic risk reduction programs, comparing resulting benefits with costs of seismic retrofitting schemes. It has been verified that, if the retrofit cost corresponding to about 7 years of the global expenditure normally granted by the Government for the reconstruction was applied to move the dwellings of the first 200 municipalities at risk by one vulnerability class, the achieved risk reduction would be of about 57%. Future works, besides the requirement of bringing up to date the hazard assessment, need to revise the fragility curves with updated shakemaps and the new site amplification factors, to improve the building inventory, and to include the uncertainty of the amplification effects and fragility curves.

Acknowledgements

The authors would like to thank Annalisa Rosti, Andrea Penna, Maria Rota and Gianmarco De Felice for the fruitful discussions that definitely help to improve the quality of the paper. Gabriele Fiorentino has received funding from the European Union’s Horizon 2020 research program under the Marie Curie grant agreement No. 892454 (https://cordis.europa.eu/project/id/892454/it).

Funding

European Union’s Horizon 2020 research program under the Marie Curie grant agreement No.

892454 (https://cordis.europa.eu/project/id/892454/it).

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Albarello D., Bosi V., Bramerini F., Lucantoni A., Naso G., Peruzza L., Rebez A., Sabetta F. & Slejko D. (2000) Carte di pericolosità sismica del territorio nazionale, Quaderni di geofisica, 12 (in Italian).
Amanti, M., et al. (2008) The Lithological map of Italy at 1: 100.000 scale: An example of re-use of an existing paper geological map. 33rd International Geological Conference, IEI02310L–6-14th August, Oslo (Norway). 2008.
Asprone, D., Jalayer, F., Simonelli, S., Acconcia, A., Prota, A., & Manfredi, G. (2013) Seismic insurance model for the Italian residential building stock. Structural safety, 44:70-79.
Borzi, B., Crowley, H., & Pinho, R. (2008) Simplified pushover-based earthquake loss assessment (SP-BELA) method for masonry buildings. International Journal of Architectural Heritage, 2(4): 353-376.
Bramerini F, Di Pasquale G (2008) Updated seismic risk maps for Italy. Ing Sismica XXV (2):5–23 (in Italian).
Camera dei deputati (2009) I principali eventi sismici a partire dal 1968. http://documenti.camera.it/leg16/dossier/Testi/Am0065.htm
Cornell, C. A. (1968) Engineering seismic risk analysis, Bull. Seismol. Soc.Am. 58, 1583–1606.
Crowley H., Colombi M., Borzi B., Faravelli M., Onida M., Lopez M., Polli D., Meroni F., Pinho R. (2009) A comparison of seismic risk maps for Italy. Bulletin of Earthquake Engineering, 7(1):149–180.
Dolce M., Sabetta F., Colozza R. (1996). Seismic risk assessment in the historical center of Rome, 11th World Conference on Earthquake Engineeering, paper n. 298 – CD-ROM.
Dolce M., Speranza E., Giordano F., Borzi B., Bocchi F., Conte C., Di Meo A., Faravelli M., Pascale V. (2019) - Observed damage database of past italian earthquakes: the Da.D.O. WebGIS. Bollettino di Geofisica Teorica e Applicata, 60(2):141-164.
Dolce M., Prota A., Borzi B., Da Porto F., Lagomarsino S., Magenes G., Moroni C., Penna A., Polese M., Speranza E., Verderame GM., Zuccaro G. (2020) Seismic risk assessment of residential buildings in Italy. Bulletin of Earthquake Engineering, 19:2999–3032.
Di Ludovico M., Prota A., Moroni C., Manfredi G., Dolce M. (2017) Reconstruction process of damaged residential buildings outside historical centres after the L’Aquila earthquake: part I “light damage” reconstruction, part II: “heavy damage” reconstruction. Bulletin of Earthquake Engineering, 15(2):667–729.
Di Pasquale G., D., Orsini G., & Romeo R. W. (2005) New developments in seismic risk assessment in Italy. Bulletin of Earthquake Engineering, 3(1):101-128.
DPC (2018) Dipartimento della Protezione Civile, Commissione tecnica per il supporto e il monitoraggio degli studi di Microzonazione Sismica (ex art.5, OPCM3907/10), www.webms.it.
Falcone G., Acunzo G., Mendicelli A., Mori F., Naso G., Peronace E., Porchia A., Romagnoli G., Tarquini E., Moscatelli M. (2021) Seismic amplification maps of Italy based on site-specific microzonation dataset and one-dimensional numerical approach. Engineering Geology 289:1-25. https://doi.org/10.1016/j.enggeo.2021.106170
GEM (2019) Italy Risk Profile, downloadable at https://downloads.openquake. org/countryprofiles/ITA.pdf
GNDT- ING-SSN (1996) Risultati dell’analisi di rischio sismico per tutta la nazione riferita al patrimonio abitativo. Elaborazioni basate sui dati ISTAT 1991. Roma (in Italian)
Goda K., Wenzel F., & Daniell J. (2014) Insurance and reinsurance models for earthquake. In Encyclopedia of earthquake engineering (pp. 1-24). Berlin: Springer.
Guidoboni E., Ferrari G., Tarabusi G., Sgattoni G., Comastri A., Mariotti D., Ciuccarelli C., Bianchi M.G., Valensise G. (2019) CFTI5Med, the new release of the catalogue of strong earthquakes in Italy and in the Mediterranean area, Scientific Data 6, Article number: 80 (2019). doi: https://doi.org/10.1038/s41597-019-0091-9
Iwahashi, J., Kamiya, I., Matsuoka, M., & Yamazaki, D. (2018) Global terrain classification using 280 m DEMs: segmentation, clustering, and reclassification. Progress in Earth and Planetary Science, 5(1):1-31.
ISTAT (National Institute of Statistics) (2001) 14° Censimento generale della popolazione e delle abitazioni, released by ISTAT on December 9, 2004, download from dawinci.istat.it (in Italian).
ISTAT (National Institute of Statistics) (2011) 15° Censimento generale della popolazione, https://www.istat.it/it/censimenti-permanenti/censimenti-precedenti/popolazione-e-abitazioni/popolazione-2011. (in Italian)
Lucantoni A., Bosi V., Bramerini F., De Marco R., Lo Presti T., Naso G., Sabetta F. (2001) Seismic risk in Italy. Ing Sismica XVI I (1):5–36 (in Italian)
Meletti C., Montaldo V. (2007). Stime di pericolosità sismica per diverse probabilità di superamento in 50 anni: valori di ag. Progetto DPC-INGV S1, Deliverable D2, http://esse1.mi.ingv.it/d2.html.
Mendicelli A., Falcone G., Acunzo G., Mori F., Naso G., Peronace E., Porchia A., Romagnoli G., Tarquini E., Moscatelli M. (2022) Italian seismic amplification factors for peak ground acceleration and peak ground velocity. Journal of Maps, pp. 1-11.
https://doi.org/10.1080/17445647.2022.2101947.
Michelini A., Faenza L., Lauciani V. & Malagnini L. (2008). ShakeMap implementation in Italy. Seismological Research Letters, 79(5):688-697.
Mori F., Mendicelli A., Moscatelli M., Romagnoli G., Peronace E., Naso G. (2020) A new VS30 map for Italy based on the seismic microzonatio dataset. Engineering Geology 275:1-10. https://doi.org/10.1016/j.enggeo.2020.105745;
NTC (2018) Norme Tecniche per le Costruzioni. Decreto Ministero Infrastrutture. GU Serie Generale n.42 del 20-02-2018 - Suppl. (Ordinario n. 8).
Rosti, A., Rota, M., & Penna, A. (2020a) Empirical fragility curves for Italian URM buildings. Bulletin of Earthquake Engineering, 19(8):3057-3076.
Rosti, A., Del Gaudio, C., Rota, M., Ricci, P., Di Ludovico, M., Penna, A., & Verderame, G. M. (2020b) Empirical fragility curves for Italian residential RC buildings. Bulletin of Earthquake Engineering, 19(8):3165-3183.
Rota, M., Penna, A., Strobbia, C., & Magenes, G. (2011) Typological seismic risk maps for Italy. Earthquake Spectra, 27(3):907-926.
Rovida A., Camassi R., Gasperini P. e M. Stucchi (2011) Catalogo Parametrico dei Terremoti Italiani (CPTI11). Istituto Nazionale di Geofisica e Vulcanologia (INGV), Milano, Bologna. https://doi.org/10.6092/INGV.IT-CPTI11
Sabetta, F., Goretti, A., & Lucantoni, A. (1998) Empirical fragility curves from damage surveys and estimated strong ground motion. Proceedings of the 11th European conference on earthquake engineering, Paris, France (pp. 1-11).
Senato della Repubblica (2017) Terremoti; L'Aquila, Reggio-Emilia, Centro Italia: politiche e risorse per ricostruire il Paese. https://www.senato.it/application/xmanager/projects/leg17/attachments/documento/files/000/028/605/Dossier_Terremoti.pdf
Silva V., Amo-Oduro D., Calderon A., Costa C., Dabbeek J., Despotaki V., & Pittore, M. (2020) Development of a global seismic risk model. Earthquake Spectra, 36 (1_suppl):372-394.
Stucchi M., Akinci A., Faccioli E., Gasperini P., Malagnini L., Meletti C., Montaldo V., Valensise G. (2004). Mappa di Pericolosità sismica del territorio Nazionale http://zonesismiche.mi.ingv.it/documenti/rapporto_conclusivo.pdf (in Italian).
Stucchi M., Meletti C., Montaldo V., Crowley H., Calvi GM., Boschi E. (2011) Seismic hazard assessment (2003-2009) for the Italian building code. Bull. Seism. Soc. of Am. 101:1885–1911.
Valensise, G., Tarabusi, G., Guidoboni, E., & Ferrari, G. (2017). The forgotten vulnerability: A geology-and history-based approach for ranking the seismic risk of earthquake-prone communities of the Italian Apennines. International journal of disaster risk reduction, 25:289-300.
Wald, D. J., and T. I. Allen (2007). Topographic slope as a proxy for seismic site conditions and amplification, Bull. Seism. Soc. Am., 97(5):1379-1395.
Zanini M.A., Hofer L., Pellegrino C. (2019) A framework for assessing the seismic risk map of Italy and developing a sustainable risk reduction program. Int J Disaster Risk Red 33:74–93
Zuccaro G. (2004) Inventory and vulnerability of the residential building stock at a national level, seismic risk and social/economic loss maps. SAVE, CD-ROM, Naples, Italy, INGV/GNDT-Project (in Italian)
Zuccaro G., Cacace F. (2011) Seismic casualty evaluation: the Italian model, an application to the L’Aquila 2009 event. In: Spence R., So E., Scawthorn C. (eds). Human casualties in earthquakes. Springer, Netherlands, pp 171–184

Download PDF

Journal Publication

published 03 Feb, 2023

Read the published version in Bulletin of Earthquake Engineering →

Reviewers agreed at journal
19 Sep, 2022
Reviewers invited by journal
18 Sep, 2022
Editor invited by journal
18 Sep, 2022
Editor assigned by journal
16 Sep, 2022
First submitted to journal
13 Sep, 2022

You are reading this latest preprint version

Influence of local site effects on seismic risk maps and ranking of Italian municipalities

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Methodology

3. Hazard And Site Effects

4. Exposure: Building Inventory And Vulnerability Classes

5. Fragility Curves

6. Seismic Risk And Damage States

7. Impact Indicators

8. Comparison With Historical Earthquakes

9. Economic Losses

10. Risk Index For Prevention Planning

11. Conclusions

Declarations

References

Status:

Journal Publication

Version 1