You are on page 1of 15

Engineering Geology 297 (2022) 106506

Contents lists available at ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

New insights on bedrock morphology and local seismic amplification of the


Castelnuovo village (L’Aquila Basin, Central Italy)
Marco Spadi a, *, Marco Tallini a, Matteo Albano b, Domenico Cosentino c, Marco Nocentini d,
Michele Saroli e, b
a
Dipartimento di Ingegneria Civile, Edile-Architettura e Ambientale (DICEAA), Università degli Studi dell’Aquila, Via Giovanni Gronchi, 18, 67100 L’Aquila, Italy
b
Istituto Nazionale di Geofisica e Vulcanologia (INGV), via di Vigna Murata 605, 00143 Roma, Italy
c
Dipartimento di Scienze, Università degli Studi Roma Tre, Largo San Leonardo Murialdo, 1, 00146 Roma, Italy
d
Istituto Superiore per la Protezione e la Ricerca Ambientale - ISPRA, Via Vitaliano Brancati, 48, 00144, Roma, Italy
e
Dipartimento di Ingegneria Civile e Meccanica (DICeM), Università degli Studi di Cassino e del Lazio meridionale, via G. di Biasio 43, 03043 Cassino, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: The Castelnuovo village is placed on a small NW-SE trending ridge, approximately 60 m higher than the valley
Castelnuovo hill floor, occupying a portion of the larger continental L’Aquila Basin (Central Italy). During the April 6, 2009
Deep boreholes L’Aquila earthquake (Mw 6.3), the village suffered heavy damage. Several studies investigated the local seismic
Microtremor measurements
amplification of the Castelnuovo area employing geotechnical, geophysical, and geological surveys, together
Resonant frequency
Impedance contrast
with 1D, 2D and 3D numerical models. However, all these studies relied on shallow geotechnical and geophysical
Geostatistical analysis surveys, which do not reach the engineering bedrock and do not constrain the presence of an impedance contrast
at depth. To date, no detailed study has been carried out to assess the depth of the engineering bedrock. In this
work, we fill this gap by executing two deep boreholes reaching the engineering bedrock, tied with an extensive
campaign of microtremor measurements all over the Castelnuovo ridge and the surrounding plain. The inter­
pretation of such new data, together with analytical, numerical, and geostatistical techniques, demonstrates that
local seismic amplification is linked to a strong impedance contrast at more than 200-m depth beneath the
Castelnuovo village associated with the lithological transition between clayey silts and breccias. Such results
differ from those provided by previous studies, where such impedance contrast was considered shallower, and
represent a milestone for assessing the local seismic hazard of the area.

1. Introduction 2013; Tallini et al., 2012, 2020).


Most of these features characterize the Castelnuovo area, located 20
The Italian Apennines (central Italy) are characterized by remarkable km ESE of L’Aquila and placed within the tectonically active Paganica-S.
seismicity and high seismic hazard, as testified by the moderate to strong Demetrio-Castelnuovo Basin (PSC) Basin, the latter considered as part of
magnitude earthquakes that dramatically devastated the area in his­ larger L’Aquila Basin (Giaccio et al., 2012; Nocentini et al., 2018)
torical and recent times (Chiarabba et al., 2005; Rovida et al., 2016; (Fig. 1a). The area owes its name to the namesake village and represents
Guidoboni et al., 2012). In this context, the amplitude and frequency a peculiar case study due to the presence of stratigraphic, topographic
content of the seismic waves are modulated by sharp changes in stiffness and morphological heterogeneities, and the presence of a network of
and thickness of soft sediments overlying the engineering bedrock shallow cavities below the buildings. The Castelnuovo village rises on a
(stratigraphic effect), by morphological features, such as steep slopes small NW-SE trending ridge with slopes up to 30 degrees, whose hilltop
(topographic effects), by the lithological heterogeneity of the sedimen­ (at approximately 860 m a.s.l.) stands of about 60 m with respect to the
tary bodies, and by the presence of potential seismic instabilities, such as surrounding plain (Fig. 1b). The village was heavily damaged during the
active faults, liquefaction prone-sites, landslides, and underground April 6, 2009, Mw 6.3 L’Aquila earthquake (MCS intensity: 9–10)
cavities (Amanti et al., 2020; Antonielli et al., 2020; Lai et al., 2020; (Chiarabba et al., 2009) and by past earthquakes from XVI century till
Lanzo et al., 2011b; Martelli et al., 2012; Saroli et al., 2014; Storti et al., now, e.g., the Mw 7.2 central-southern Apennines (1456) and the Mw 7.1

* Corresponding author.
E-mail address: marco.spadi@univaq.it (M. Spadi).

https://doi.org/10.1016/j.enggeo.2021.106506
Received 26 February 2021; Received in revised form 14 December 2021; Accepted 21 December 2021
Available online 27 December 2021
0013-7952/© 2021 Elsevier B.V. All rights reserved.
M. Spadi et al. Engineering Geology 297 (2022) 106506

Marsica earthquake (1915). The MCS intensity values estimated for each Castelnuovo area employing shallow geotechnical, geophysical, and
of these seismic events was higher than eight, independently from geological surveys, coupled with 1D, 2D and 3D numerical models. Such
earthquake magnitude and epicentral distance (Galli et al., 2011; Rovida studies provided different interpretations about the factors responsible
et al., 2016). for the village’s heavy damage after the L’Aquila 2009 earthquake.
Previous studies investigated the local seismic amplification of the Ameri et al. (2011) exploited S-wave spectral amplitudes from 112

Fig. 1. a) Geo-lithological map of the Castelnuovo area, modified from Nocentini et al. (2018). The upper-right inset shows the Italian seismic hazard map, with peak
ground accelerations (g) with a 10% chance of being exceeded in 50 years (Meletti and Montaldo, 2007). The lower-left inset shows the stratigraphic scheme
illustrating the formations distribution around Castelnuovo village. Key to the legend: ALL = Alluvial deposits (Upper Pleistocene - Holocene). COL = Colluvial
deposits (Upper Pleistocene - Holocene). CON = Conglomerates deposits (Valle dell’Inferno and Valle Orsa formations) (Piacenzian-Gelasian). SI = Clayey silts (San
Nicandro Fm.) (Piacenzian-Gelasian). BR = Breccia deposits (Madonna della Neve and Valle Valiano formations) (Piacenzian-Gelasian). CB = Carbonate bedrock
(Meso-Cenozoic). 1 = Detail of the study area in Fig. 2. 2 = Inferred normal fault. 3 = Traces of cross-sections in Fig. 10. 4 = Buildings. b) 3D view of Castelnuovo
village from southeast.

2
M. Spadi et al. Engineering Geology 297 (2022) 106506

aftershocks registered during the L’Aquila 2009 seismic sequence and deposited in a lacustrine environment (Spadi et al., 2016). SI deposits in
concluded that amplifications in the Castelnuovo village were due to the Castelnuovo hill contain several anthropic cavities, which are
topographic effects. Lanzo et al. (2011a) carried out 1D and 2D dynamic approximately 25to 30 m long, 2 to 6 m high and are located 2–10 m
numerical simulations of the Castelnuovo hill and concluded that the below ground level. Above or heteropic to SI, a Gilbert-type delta system
seismic response of the Castelnuovo hill appeared predominantly gradually formed, with deposition sandy gravel deposits of Valle Orsa
affected by topographic amplification. Sica et al. (2014) investigated the Formation and subsequent Valle dell’ Inferno Formation composed by
role exerted by underground cavities on the ground motion at the sur­ sub-horizontal beds of clast-supported conglomerates belonging to a
face of the Castelnuovo hill through 2D nonlinear numerical analyses. gravel-bed braided fluvial system (CON in Fig. 1a) (Giaccio et al., 2012;
They concluded that the presence of multiple shallow cavities had a non- Cosentino et al., 2019).
negligible effect on the local seismic amplification. Finally, Landolfi The younger synthems of the PSC Basin unconformably overlay the
(2013) and later Evangelista et al. (2016) performed a 3D dynamic San Demetrio-Colle Cantaro Synthem and were deposited in smaller sub-
numerical model constructed by interpreting the 2D cross-sections of basins all along the entire PSC Basin, with western migrated depocenters
Lanzo et al. (2011a) and declared that the topographic effects signifi­ (Giaccio et al., 2012; Tallini et al., 2019). Nearby the Castelnuovo
cantly influenced the ground motion at the surface. In contrast, the role village, these more recent deposits occupy the lower part of the valley
of cavities seemed to be negligible. floor with colluvial and alluvial deposit (ALL and COL in Fig. 1a), lying
All of these studies and the performed numerical models were cali­ in areas lower than 800 m a.s.l., and generically attributed to the Upper
brated interpreting shallow geotechnical surveys and geophysical mea­ Pleistocene-Holocene (Nocentini et al., 2018).
surements (less than 50 m depth) only, performed mainly on the
Castelnuovo hilltop and not reaching the engineering bedrock. Other­ 3. Data and methods
wise, no deep boreholes were executed in the area to investigate the
depth of the engineering bedrock and to assess the presence of an The investigation of the subsoil of the Castelnuovo hill (the area
impedance contrast at depth. enclosed within the dashed black rectangle in Fig. 1a) and its local
Assessing the seismic amplification at local scale requires developing seismic amplification has been conducted according to the following
2D or 3D numerical models that must be sufficiently constrained with steps:
deep geotechnical and geophysical measurements to provide a more
detailed picture of the bedrock shape and the soil linear and nonlinear 1. Collection and interpretation of data available from literature, con­
dynamic properties. Such an aspect is fundamental since basin geometry sisting in geological maps, shallow boreholes and geophysical sur­
and heterogeneity strongly affect the propagation of seismic waves and veys, in situ and laboratory geotechnical tests, and shallow Vs-
their amplitude and frequency content (Chandran and Anbazhagan, profiles (less than 50 m).
2017; Saroli et al., 2020). In this work, we provide an up-to-date picture 2. Execution of new deep boreholes, extending at depths higher than
of the subsoil beneath the Castelnuovo hill and the depth of the engi­ 100 m and reaching the BR lithology (Fig. 1a), together with an
neering bedrock by performing and interpreting for the first time the extensive campaign of single station microtremor measurements all
findings of a deep borehole drilled at the top of the Castelnuovo hill up to over the Castelnuovo hill and the surrounding plain.
220 m depth and reaching the engineering bedrock. We also provide 3. Spatial analysis of resonant frequencies employing a geostatistical
new insights about the local fundamental frequencies of oscillation of approach.
the subsoil by performing an extensive campaign of single station 4. Assessment of the impedance contrast at depth through 1D dynamic
microtremor measurements all over the Castelnuovo hill and the sur­ modeling.
rounding plain. All these data were interpreted through geostatistical
and numerical approaches and confirmed the presence of a deep
impedance contrast, with resonant frequencies in the range of 0.7–3.8 3.1. Available data from the literature
Hz. The latter is attributed to a lithological passage between the clayey
silts of the San Nicandro Formation and the breccias of Madonna Della Available in situ and laboratory geotechnical surveys consist of
Neve Formation, as highlighted by the performed deep borehole. Such a shallow boreholes (BDH1 to BDH6 the blue triangles in Fig. 2), NSPT
new finding points out that local seismic amplification of the area is blow counts, grain size distributions, consistency limits and soil unit
mainly modulated by the stratigraphic effect, thus providing an updated weight measurements (Ilic et al., 2017; Working Group MS–AQ, 2010).
description of the subsoil’s linear dynamic properties with respect to Geophysical measurements consist of seven Vs versus depth profiles
previous studies, helpful in assessing the seismic hazard of the area. obtained from the interpretation of six down-hole tests performed in the
boreholes BDH1 to BDH6 (the blue triangles in Fig. 2) and one MASW
2. Geological setting (the green square in Fig. 2) (Working Group MS–AQ, 2010). Such Vs-
profiles were validated by estimating Vs-values with depth with NSPT
The Paganica-San Demetrio-Castelnuovo (PSC) Basin is an inter­ using empirical relations proposed by the literature (Bajaj and Anbaz­
montane sedimentary basin filled with a thick succession of Plio- hagan, 2019).
Quaternary continental deposits, which unconformably overly the
Meso-Cenozoic bedrock (CB in Fig. 1a). The latter consists of platform- 3.2. New boreholes
and-slope limestones stacked onto upper Miocene turbidites during
the Apennine late Messinian-Zanclean compressional tectonics (Cosen­ A 223 m-depth borehole located at the top of the Castelnuovo hill
tino et al., 2010). The post-orogenic succession is divided into several (the red diamond CN1 in Fig. 2 and supplementary Fig. S1a) and a core
synthems; the oldest one is called San Demetrio-Colle Cantaro Synthem, destruction drilling borehole, 50 m deep, located to the east of the
which comprises several formations deposited in different environ­ Castelnuovo village (the red diamond CN2 in Fig. 2 and supplementary
ments, mainly related to a lacustrine system (Cosentino et al., 2017; Fig. S1b) were drilled to investigate the thickness of the SI lithology
Nocentini et al., 2018). Along the northeastern margin of the area, al­ beneath Castelnuovo hill. The CN1 borehole was carried out using a
luvial fan and slope deposits developed with alluvial breccia deposits of continuous drilling EGT 710 MD probe. The drilling was performed
Valle Valiano Formation and intercalations of silts and fan of Madonna using a 1 m long core barrel with a diameter of 101 mm. The lining of the
Della Neve Formation, here grouped in the Breccia deposits unit (BR in hole was made with metal pipes 1.50 m long and with a diameter of 127
Fig. 1a). The lacustrine deposition is represented by San Nicandro For­ mm, while the drilling rods have a diameter of 89.9 mm and lengths of
mation (SI in Fig. 1a), which consists of massive whitish calcareous silts 1.50 and 3.00 m. The extracted samples were stored in PVC cataloguing

3
M. Spadi et al. Engineering Geology 297 (2022) 106506

Fig. 2. Detail of the area enclosed in the dashed black rectangle in Fig. 1a showing the lithological map of the site, with the location of the available and new data.
Key to the legend: 1) New boreholes. 2) Available shallow boreholes and downhole. 3) MASW. 4) Microtremor measurements. 5) Traces of cross-sections in Fig. 10. 6)
Inferred normal fault. For the lithological symbols, refer to the legend in Fig. 1.

boxes. velocity bedrock according to the horizontal-to-vertical spectral ratio


The CN2 deep borehole was performed with the core-destruction method (hereinafter HVSR) (Nakamura, 1989, 2019). HVSR method has
drilling technique (see Fig. S1b in supplementary material). been exploited for the estimation of site resonant frequency and inverted
We followed the ISO 14688:2017 and ASTM D2487–11 standards for for the retrieval of shear-wave velocity profiles (Vs) in several seismic
the borehole execution, soil logging and sampling. The stratigraphic logs microzonation and dynamic site characterization studies (Del Monaco
of these two boreholes were interpreted with the contribution of et al., 2013; Gosar and Lenart, 2010; Parolai, 2012; Saroli et al., 2020).
detailed geological outcropping data and the 1:25000 geological map of This method splits the horizontal and vertical components of a micro­
Nocentini et al. (2018). tremor measurement into several time windows of equal or varying
length and removes windows affected by strong noise transients. Then,
smoothed Fourier spectra are computed, and the ratio between the
3.3. Implementation and processing of microtremor measurements horizontal spectra, given by the mean of the two horizontal components,
and the vertical spectra, is performed for each time window. Finally, the
An extensive campaign of single station microtremor measurements mean HVSR curve is calculated by averaging the horizontal-to-vertical
has also been performed at 71 sites over the Castelnuovo hill and the spectral ratios of each time window (Molnar et al., 2018).
surrounding areas to identify peak frequencies potentially associated The HVSR curve is estimated with the Geopsy code (www.geopsy.org
with deep impedance contrasts (pink circles in Fig. 2). ) (Wathelet et al., 2020) for each available microtremor measurement.
The spatial distribution of measurements is irregular. Such uneven Amplitude spectra were computed in the frequency interval of 0.25–20
distribution is due to logistic impediments related to densely vegetated Hz. The three spatial components of the recorded measurement were
areas and private properties. divided into non-overlapping windows of 40 s long to guarantee at least
We used a three-component SL07 seismometer by S.A.R.A, equipped ten selected windows, according to the SESAME (2004) criteria. Win­
with an internal 2 Hz sensor (http://www.sara.pg.it) and set to a sam­ dows affected by transients were automatically removed based on an
pling frequency of 100 Hz (supplementary Fig. S1c). We adopted a anti-trigger algorithm that performs a comparison between the short-
recording length of approximately 30 min for each measurement, which term average (STA) and the long-term average (LTA) level of signal
is recommended for spectral analysis at least down to 0.5 Hz (SESAME, amplitude, fixed to 1 s and 30 s, respectively. Only the windows with an
2004). The effect of anthropic signals has been mitigated by performing STA/LTA ratio between 0.1 and 2.5 were kept for the HVSR computation
recordings at different times during the day and night and more than 10 (SESAME, 2004).
m far from infrastructures and trees (Molnar et al., 2018; Panzera et al., Finally, the computed Fourier spectra were smoothed with the
2019). The recordings at a single site were repeated in case of in­ Konno–Ohmachi function (Konno and Ohmachi, 1998), with a band­
consistencies with the neighboring measurements. width coefficient b = 40. Such parameters are commonly adopted to
Microtremor measurements were exploited to evaluate the funda­ process microtremor measurements in Italy (Maresca and Berrino, 2016;
mental frequency (f0) of unconsolidated sediments lying over high-

4
M. Spadi et al. Engineering Geology 297 (2022) 106506

Gallipoli et al., 2020; Lanzo et al., 2011a).

3.4. Spatial analysis of resonant frequencies

The spatial relationship between the resonant frequencies has been


investigated using a geostatistical approach. Such an approach has been
successfully applied to (i) identify one or more clusters of frequency
peaks related to different impedance contrasts at depth, (ii) assess the
presence of trends in the measurement distributions, (iii) identify sharp
peaks or shifts in the frequency values, potentially associated to the
presence of buried faults at depth, (iv) assess if there is a spatial corre­
lation between the obtained measurement peaks, and (v) provide the
uncertainty associated to the interpolated map (Saroli et al., 2020;
Trevisani et al., 2017, 2021). In this study, we exploited Kriging inter­
polation techniques (Krige, 1951), including a variety of least-squares
methods that provide predictions with the minimum variance of the
investigated variable (Oliver and Webster, 2014). First, a preliminary
exploratory spatial analysis of f0 has been conducted to recognize
possible correlations, trends, and outliers. Then, the experimental var­
iogram has been estimated to identify the presence of possible anisot­
ropies in the data. Finally, the experimental variogram has been fitted
with a gaussian plus nugget model defined by Eq. (1) (Oliver and
Webster, 2014):
⎧ { ( 2) }
⎨ c + c 1 − exp − h

for 0 < h
(1)
0
γ(h) = a2


0 for h = 0

where h is the lag distance, c0 is the nugget, c0 + c is the sill, and a is a


distance parameter, proportional to the effective range (r) according to
√̅̅̅̅̅̅
equation r = 3a. Experimental data have been interpolated with or­
dinary kriging, together with the associated variance (or standard
deviation).
The complete analysis has been performed with the geostatistical
package Gstat (Pebesma, 2004), accessible through the open-source R
code (R Team Core, 2019), together with the open-source GIS environ­
ment QGIS (QGIS Development Team, 2018).
Fig. 3. (a) Typical stratigraphy at BDH1 to BDH6 boreholes. (b) Soil grain size
3.5. Assessment of the presence of an impedance contrast at depth distributions from available laboratory tests. (c) Trend of the soil natural unit
weight with depth (d) Shear-wave velocity profiles from available Down-hole,
MASW and interpreted from NSPT blow counts (Ilic et al., 2017; Working
A 1D dynamic numerical model has been constructed with the
Group MS–AQ, 2010).
STRATA code (Kottke et al., 2019). Available borehole logs, Vs-profiles
from down-hole and MASW measurements, and laboratory data have
been exploited to simulate the propagation of the seismic wave in the 4. Results
study area and to assess the presence of an impedance contrast at depth.
We assumed a 1D soil column made of an elastic isotropic material, 4.1. Available geotechnical properties of the Castelnuovo hill subsoil
neglecting the nonlinear behavior of soils. The stiffness variation with
depth has been assumed considering a first-order average velocity-depth The available shallow borehole logs and grain size distributions
function according to the empirical Eq. (2) (Ibs-Von Seht and Wohlen­ testify that the soil in the first 50 m depth is almost homogeneous. It is
berg, 1999; D’Amico et al., 2008): composed of clayey silts with centimetric to decametric levels of sand
( ) belonging to the San Nicandro Formation (Fig. 3a-b and SI in Fig. 2),
with average water content and degree of saturation of 44.9% and 96%,
Vs (z) = Vs0 1 + z/z x (2)
0 respectively (Ilic et al., 2017). The average natural unit weight is
approximately 17 kN/m3 and does not show any dependency with
where z is the depth below the ground, z0 = 1 m, Vs0 is the shear wave depth, at least in the first 30 m (Fig. 3c). Mean Atterberg limits of
velocity at the surface, and x is an exponential term defining the depth approximately 58.4% for the liquid limit and 38.4% for the plastic limit
dependence of the shear-wave velocity. Eq. (2) coefficients (Vs0 and x) classify the SI lithology as moderately plastic (plasticity index =10) and
are calibrated as the Vs-profile falls in the range and trend defined by with solid consistency (consistency index = 1.35) (Ilic et al., 2017;
experimental data in Fig. 3d. The obtained 1D transfer function is then Working Group MS–AQ, 2010).
compared with the experimental H/V curves from microtremor mea­ The available Vs profiles with depth (Fig. 3d), obtained from
surements. Such a comparison shows a strong fit by Oubaiche et al. Downhole and MASW, show values between 180 m/s and 500 m/s with
(2016) and has been applied in several other studies (Saroli et al., 2020; an overall increase with depth. Such trend is also confirmed by inter­
Di Naccio et al., 2020; Lanzo et al., 2011a). preting the NSPT blow counts executed at different depths in boreholes
BDH1-BDH3 with the empirical relations proposed by Seed and Idriss
(1981), Ohsaki and Iwasaki (1973) and Pitilakis et al. (1999) and valid
for slits and alluvial soils. The derived Vs values from the three relations

5
M. Spadi et al. Engineering Geology 297 (2022) 106506

are averaged at different depths and plotted as green circles in Fig. 3d. grained deposits roughly disappear below the depth of 220 m, passing
to coarse calcareous breccias pertaining to Madonna Della Neve For­
4.2. Results of the new borehole logs mation (BR) (Fig. 4h). The CN1 borehole represents the first direct ev­
idence of the considerable depth of the Plio-Quaternary deposits in the
Two new boreholes were drilled on the top of the Castelnuovo hill PSC Basin. It also indicates for the first time the presence of a thick layer
and in the nearby plain (the red diamonds in Fig. 2) to investigate the of breccias between the lacustrine deposits and the pre-orogenic
thickness of the SI lithology and the relationship between the conti­ substratum.
nental deposits and the carbonate bedrock. The CN1 continuous drilling The CN2 destruction-core borehole is located to the east with respect
borehole (Fig. 4) is located at the hilltop at approximately 860 m a.s.l. to the CN1 borehole at the foot of the Castelnuovo hill (approximately
and shows a lithostratigraphic sequence composed mainly by alterna­ 800 m a.s.l.). It shows a similar stratigraphy but with different thick­
tions of fine-grained laminated and massive silts up to the depth of 210 nesses. Indeed, the first 50 m consist of calcareous and clayey silts
m (SI) (Fig. 4) (Cosentino et al., 2019). In detail, the upper part of the belonging to the SI lithology, while the last 1 m is composed of calcar­
well-log (depth less than 120 m), partly matching the outcropping eous breccias referable to BR breccias (Fig. 4). Both boreholes highlight
succession, is characterized by the typical features of the San Nicandro that the SI lithology thickness reduces moving to the east-southeast. This
Formation, consisting of a cyclic alternation of white calcareous silts and reduction of approximately 160 m is due in part to the change in altitude
light grey clayey silts. In this interval, the CaCO3 content cyclically between the two boreholes (about 60 m; CN1 ≈ 860 m a.s.l.; CN2 ≈ 800
ranges between 50% and 90% (Fig. 4a–b). In the lower part of the well- m a.s.l.) and part to the irregular morphology of the underlying breccias,
log (depth greater than 120 m), clayey levels become abundant and whose roof increases his height of approximately 100 m moving from
thicker, and organic-rich clay levels appear. In contrast, minima CaCO3 CN1 to CN2 (Fig. 4).
contents are periodically recorded, around 25% (Fig. 4c–g). The fine-

Fig. 4. Schematic stratigraphy of the CN and CN2 borehole, together with the pictures of the main lithofacies of the CN1 borehole. a) laminated to massive
calcareous silt, t: blackish tephra, picture courtesy of Giovanni Zanchetta; b) laminated to massive greyish calcareous clayey silt with organic horizons (oh); c)
laminated to massive greyish silty clays with soft-sediment deformation (sd); d) laminated darkish organic silty clays with calcareous dropstone (d); e) laminated to
massive calcareous silt, sandy silt, and sand with a calcareous pebble; f) massive pebbly darkish clays; g) massive to laminated greyish clayey silt; h) coarse polygenic
calcareous breccia (Piacenzian breccia). a and b: San Nicandro silt; c, d, e, f, and g: Piacenzian clay.

6
M. Spadi et al. Engineering Geology 297 (2022) 106506

4.3. HVSR results and spatial mapping by the SESAME working group (SESAME, 2004). Frequency peaks not
fulfilling the SESAME criteria were discarded. Still, three measurements
Before interpreting the obtained H/V peaks according to the subsoil non-fulfilling the SESAME criteria were deemed resonant frequencies.
dynamic properties, we checked the possible occurrence of measure­ The obtained frequency peaks were consistent with to those at verified
ment issues and anomalous peaks in the HVSR curve. Some measure­ measurements located at no more than 95 m (i.e., the average nearest
ments were omitted because of recording issues related to wrong neighbor distance of the whole dataset) at the closest locations. The
positioning of the instrument, atmospheric disturbances, or anthropo­ complete set of the identified frequency peaks comprises 66 measure­
genic interference on the device itself. Some measures, repeated at night ments and is reported in Fig. 5 with some relevant H/V plots (the full list
in the exact locations, did not show abnormal frequency peaks, thus of the H/V peak frequency is available in Fig. S2 and Table S1 in the
identifying a resonant frequency at the site. For each H/V curve, one or Supplementary Material).
more resonance peaks have been assessed following the criteria defined Almost all measurements show a single resonant peak in the range

Fig. 5. Map of resonant frequencies at measurements performed over the Castelnuovo area with some examples of H/V curves. The dashed white line indicates the
contour line at 820 m a.s.l. For the other symbols, please refer to Fig. 1.

7
M. Spadi et al. Engineering Geology 297 (2022) 106506

0.7–5.8 Hz (see S1 to S30 H/V curves in Fig. 5). Eight measurements HVSR curves for f0 < 3.8 Hz show bright and sharp frequency peaks,
only show a second peak at frequencies higher than 6 Hz (see for whose shape and frequency values agree with the results of microtremor
example, the S39 H/V curve in Fig. 5). These double peaks are neither surveys performed in previous studies on the Castelnuovo hill only
clustered on a specific area nor directly associated with a very shallow (Gallipoli et al., 2011, 2013; Lanzo et al., 2011a). Based on preliminary
impedance contrast. Therefore, the subsequent analyses focused only on analysis, these resonant frequencies could be related to an impendence
the first frequency peak, the latter being observable all over the study contrast located between 30 and more than 100 m at depth (Albarello
area. and Castellaro, 2011). Such an impendence contrast could be linked to
A preliminary statistical analysis of the computed resonant fre­ the passage between the clayey silts (SI) and the gravels and breccias
quencies (Fig. 6a) highlights that the whole dataset comprises two (BR) as observed in CN1 and CN2 boreholes (Fig. 4).
distinct families, which separate at approximately 3.8 Hz. The frequencies for f0 > 3.8 Hz range between 3.8 and 5.8 Hz and
The first group for f0 < 3.8 Hz (56 values) is located over the hill and present small amplitude peaks, suggesting that the impendence contrast
in the immediate surroundings (Fig. 5), where the clayey silts (SI) and is shallow (i.e., less than 30 m) (Albarello and Castellaro, 2011) and
the alluvial/colluvial deposits (ALL and COL) mainly outcrop. The sec­ probably associated with the presence of BR over the Meso-Cenozoic
ond group for f0 > 3.8 Hz (10 values) comprises measurements located carbonate bedrock (CB in Fig. 1).
at sites to the northern and eastern margins of the study area (the dark- The spatial analysis has been conducted considering the first group of
purple circles in Fig. 5), where breccias sub crop or outcrop (BR). frequency peaks for f0 < 3.8 Hz (Fig. 6b), i.e., the measurements

Fig. 6. a) Resonant frequency distribution for all the 66 microtremor measurements. b) Resonant frequency distribution for f0 < 3.8 Hz, skewness = 2.12. c)
Scatterplot of resonant frequencies for f0 < 3.8 Hz vs longitude. d) Scatterplot of resonant frequencies for f0 < 3.8 Hz vs latitude. e) Scatterplot of resonant frequencies
for f0 < 3.8 Hz vs altitude. f) Resonant period distribution, skewness = − 0.56.

8
M. Spadi et al. Engineering Geology 297 (2022) 106506

performed all over the Castelnuovo hill and in the surrounding plain. and directional variograms (Fig. 7c) of resonant periods do not show
Such a choice stems from the observation that different f0 families evident anisotropies, local outliers, or patterns related to non-
cannot be analyzed simultaneously. The spatial analysis for f0 < 3.8 Hz stationarity. Then we calculated the isotropic experimental variogram
highlights that frequency values gradually increase moving towards the (Fig. 7d) assuming a unit lag of 95 m and a maximum length h of 800 m.
east (Fig. 6c) and south (Fig. 6d). Some sharp changes locate at longitude The selected unit lag corresponds to the average nearest neighbor dis­
≈ 387200E (Fig. 6c) and latitude ≈ 4,683,200 N (Fig. 6d), which are tance between the dataset points, while the maximum length is fixed
probably associated with the presence of buried faults at depth. A because, at large distances, the number of sample pairs decreases, and
decrease of frequency values with increasing height is also observed in the experimental variogram becomes less stable. The final isotropic
Fig. 6e for altitude less than 820 m a.s.l. (the dashed white line in Fig. 5), experimental variogram (the light blue circles in Fig. 7d) presents a
while at higher altitude frequency values are almost constant, amount­ small nugget effect, probably associated with the uncertainty inherent to
ing at approximately 0.9 Hz. microtremor measurements.
In terms of amplitude, the HVSR curves show peaks up to 10 without The experimental variogram has been fitted with the isotropic
a clear spatial correlation with distance or altitude (supplementary Gaussian plus nugget model of Eq. (1) (red curve in Fig. 7d), whose
Fig. S3). parameters, i.e., r = 647.9 m, c0 + c = 0.135 s2, and c0 = 0.0015 s2,
A better assessment of the spatial variation of resonant frequency resulted in being the best solution, with a weighted sum of squared error
values below 3.8 Hz is attained by spatial mapping with Kriging inter­ (SSE) of 1.27 × 10− 8. Fitting the experimental variogram with different
polation. However, the f0 dataset (Fig. 6b) shows a positively skewed theoretical models (e.g., exponential or power) gave higher SSE values.
statistical distribution (skewness coefficient = 2.12) with a long tail of Once we defined the theoretical variogram, we interpolated the
high f0 values. Positively or negatively skewed data inflate variances and resonant periods using block-kriging interpolation. The obtained map is
distort the variogram (Oliver and Webster, 2014). A practical solution shown in Fig. 8a, while the standard deviation error (Std) associated
has been proposed by Trevisani et al. (2017) who suggested to work with with the predicted resonant period map is shown in Fig. 8b. The reso­
resonance periods (Fig. 6f) instead of f0. Such a modification greatly nant period map is blanked on areas where the prediction variance is
reduces the data skewness from 2.12 to − 0.56, thus improving the larger than the third quartile of the variance distribution (i.e., 0.0016
geostatistical interpolation (Trevisani et al., 2017; Saroli et al., 2020). s2), corresponding to a maximum standard deviation of 0.04 s.
The calculated variogram cloud (Fig. 7a), variogram map (Fig. 7b)

Fig. 7. Geostatistical analysis of the resonant period values. a) Experimental variogram cloud. b) Experimental variogram map. c) Directional experimental var­
iograms. d) Experimental omnidirectional variogram (the light blue circles) and best fit with a gaussian plus nugget model (red line). Numbers indicate the samples
for each class. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

9
M. Spadi et al. Engineering Geology 297 (2022) 106506

Fig. 8. Results of the geostatistical analysis performed on the resonant period values. a) Interpolated map of the resonant period using block-Kriging. The circles
identify the measured resonant period from microtremor measurements. b) Kriging prediction standard deviation map of the resonant period. The black circles
identify the location of the microtremor measurements.

4.4. Assessment of the impedance contrast at depth resonant periods (or frequencies) are be related to the SI thickness
changes or, more specifically, to the presence of an impedance contrast
The resonant period map in Fig. 8a highlights an overall decrease of at the passage from the SI and BR lithologies. To verify such a hypoth­
resonant periods (corresponding to an increase in frequency) moving to esis, we estimated the first resonant frequency at CN1 and CN2 bore­
the east-southeast direction. The CN1 and CN2 boreholes (Figs. 2 and 4) holes by simulating the 1D transfer function from the bedrock to the
show that the thickness of the SI lithology reduces moving towards the outcropping soil at the CN1 and CN2 boreholes in Fig. 4. We compared
same direction. . Therefore, we argue that the observed changes of the computed transfer function with the HVSR curves of the S1 and S39

10
M. Spadi et al. Engineering Geology 297 (2022) 106506

microtremor measurements (Fig. 5), the latter being the closest micro­ defined according to the Random Vibration Theory (RVT). The calcu­
tremor measurements to the CN1 and CN2 boreholes (1 m and 5 m, lated theoretical transfer functions at CN1 and CN2 boreholes have been
respectively). compared with the experimental H/V curve of S1 and S39 measure­
For the simulation of the theoretical transfer function at CN1 and ments, respectively.
CN2 boreholes, we constructed a 1D column made of a homogeneous The results of the analysis show that the computed 1D theoretical
linear-elastic material corresponding to SI lithology, laying on an elastic transfer functions for the CN1 and CN2 profiles (the violet curves in
half-space representing the engineering bedrock, which corresponds to Fig. 9a and b) agree with the experimental H/V curves (the red curves in
the BR lithology in Fig. 4. The SI lithology is modelled as a linear elastic Fig. 9a and b). Indeed, the computed first resonant frequency peaks,
material, with a viscous damping ratio of 5% and a mean unit weight of equal to 0.92 Hz and 2.9 Hz for CN1 and CN2, respectively, almost
17 kN/m3, selected according to the experimental values in Fig. 3c and coincide with those experimentally derived at S1 and S39 microtremor
the parametrization proposed by Lanzo et al. (2011a) and Evangelista measurements, equal to 0.94 Hz and 3 Hz, respectively. . The adopted
et al. (2016). The variation of Vs (or shear stiffness) with depth is average Vs curve (the blue curve in Fig. 9a and b), whose coefficients are
defined according to Eq. (2) for both CN1 and CN2 models. The pa­ Vs0 = 240 m/s and x = 0.24 (Eq. (2)) follows the general trend of the
rameters defining the position (Vs0) and shape (x) of the curve in Eq. (2) available experimental Vs profiles in Fig. 3d, reported as black
have been selected to follow the trend identified by the available segmented lines in Fig. 9a and b.
experimental Vs profiles in Fig. 3d. The nonlinear behavior of the SI
lithology (i.e., the decay of stiffness and increase in damping with shear 5. Discussion
strains) is neglected, given the elastic nature of microtremor measure­
ments and the associated small shear strains. Two different 1D models Different interpretations about the factors affecting the local seismic
were investigated (Fig. 9a and b), where the thickness of the SI layer amplification of the Castelnuovo village have been provided in the
varies according to that identified in the CN1 and CN2 boreholes. literature (Lanzo et al., 2011a; Landolfi, 2013; Sica et al., 2014; Evan­
The STRATA code allows introducing the uncertainty associated gelista et al., 2016). Discrepancies were attributed to uncertainties in the
with the adopted mean shear wave profile with depth by applying a log- definition of the depth of seismic bedrock and the absence of a direct
normal distribution for the mean Vs value (Toro, 1995; Toro et al., evidence of an impedance contrast at depth. Indeed, previous studies
1992), with a logarithmic std. of 0.5. In this way, the code performs 200 relied on shallow geotechnical, geophysical, and geological surveys
different simulations by randomly varying Vs-profile in the range only.
comprised in the grey areas in Fig. 9a and b. For the BR lithology, rep­ Thanks to the analysis of new deep boreholes performed in the study
resenting the elastic half-space, an average shear wave velocity of 1000 area and crosscutting the SI lithology (CN1 and CN2 in Fig. 4), and new
m/s has been assumed. Such a value has been defined according to that geophysical surveys, we provided an updated picture of the subsoil
adopted in 2D and 3D numerical models by Lanzo et al. (2011b) and beneath the Castelnuovo hill and its local seismic amplification.
Evangelista et al. (2016) and based on in-situ measurements executed on The Castelnuovo village is located on a hill, whose hilltop (at
similar lithologies (Amoroso et al., 2018). The seismic input has been approximately 860 m a.s.l.) stands of about 60 m with respect to the

Fig. 9. Results of the 1D numerical modeling. From left to right in each panel: 1D numerical model scheme adopted for the computation of the elastic transfer
function at CN1 (panel a) and CN2 (panel b) boreholes; Vs profile assumed for the SI layer according to Eq. (2). (blue curve) and comparison with the experimental Vs
profiles (black lines); computed theoretical 1D transfer function (purple curve) and comparison with the experimental HVSR curve (red curve) obtained from
adjacent microtremor measurements (S1 in panel a and S39 in panel b). (For interpretation of the references to colour in this figure legend, the reader is referred to
the web version of this article.)

11
M. Spadi et al. Engineering Geology 297 (2022) 106506

surrounding plain (Fig. 1b). The hill is mainly composed of clayey silts of thick succession of clayey silts and clays (SI) (Fig. 4). Below this rela­
medium plasticity (Fig. 3) belonging to the San Nicandro Fm. (SI in tively homogeneous succession, 10 m of breccias are present. The CN2
Figs. 1a and 2), sometimes overlapped by few meters of recent colluvium borehole, drilled at the foot of the Castelnuovo hill to the east-southeast
or alluvium at the foot of the Castelnuovo hill (ALL and COL in Fig. 1a). (Fig. 2), shows a stratigraphic sequence similar to CN1(Fig. 4), with 50
The SI lithology lays on breccias deposits (BR in Fig. 4), which outcrop m of SI lithology overlying one meter of the BR deposits. Both boreholes
on the northern and eastern margins of the area. testify that the SI lithology tapers moving to the east-southeast. This
The performed CN1 and CN2 boreholes (Fig. 4) allowed for the first tapering of approximately 160 m is due to the change in altitude be­
time to measure the thickness of the SI lithology in the area and identify tween the two boreholes (about 60 m; CN1 ≈ 860 m a.s.l.; CN2 ≈ 800 m
the BR lithology at its bottom. Previous studies inferred such informa­ a.s.l.) and to the irregular morphology of the underlying breccias, whose
tion by interpreting shallow geotechnical and geophysical surveys only, roof increases his height of approximately 100 m moving to the east-
not reaching the BR lithology and not linked to any deep borehole southeast from CN1 to CN2 (Fig. 4).
reaching the engineering bedrock (Lanzo et al., 2011a; Landolfi, 2013; The performed microtremor surveys identified clear and sharp
Sica et al., 2014; Evangelista et al., 2016). The CN1 borehole, drilled at resonant frequency peaks at measurements performed mainly on the SI
the top of Castelnuovo hill (Fig. 2), highlights the presence of a 210 m- lithology (for f0 < 3.8 Hz), in the central part of the study area (Fig. 5).

Fig. 10. Geological cross-sections of the Castelnuovo area, crossing the CN1 borehole located at the hilltop. The CN2 borehole in the AB cross-section is projected.
The cross-section traces are reported in Figs. 1a, 2 and 8a. For the symbol legend, please refer to the caption of Fig. 1a.

12
M. Spadi et al. Engineering Geology 297 (2022) 106506

The computed resonant period map (Fig. 8a) highlights that resonant Indeed, the only two boreholes reaching the engineering bedrock
period decreases (or resonant frequency increases), moving to the and the clustering of the microtremor measurements close to the
southwest, from the CN1 borehole where the resonant period is inhabited areas of the Castelnuovo village do not allow to provide a
approximately 1.1 s (Fig. 8a; resonant frequency ≈ 0.9 Hz), to the CN2 detailed picture of the engineering bedrock and the possible presence of
borehole where the resonant period is approximately 0.33 s (Fig. 8a; sharp discontinuities due to buried faults. The latter are inferred in
resonant frequency ≈ 3 Hz). We interpret these frequency peaks as the Fig. 10 according to the rate changes of the resonant period only.
resonant frequency of the SI lithology, associated with an impedance Moreover, the estimated mean Vs-depth relationship (blue curve in
contrast at depth at the transition between SI and BR lithologies. We also Fig. 9) represents a 1st-order estimation of the shear stiffness of SI li­
explain such a period (or frequency) variation to the southeast direction thology at shallow depth (less than 30 m). In contrast, at greater depth,
with the progressive tapering of the SI lithology, highlighted by the CN1 shear stiffness is constrained by the results of 1D numerical models only.
and CN2 boreholes (Fig. 4). This interpretation also explains the Currently, logistic impediments such as the presence of private proper­
apparent inverse proportionality between the f0 values and the altitude ties and highly vegetated areas, together with strict regulations, do not
in Fig. 6e. Measurements performed at heights less than 820 m a.s.l. are allow to perform an extensive campaign of boreholes reaching the en­
distributed over a large area whose topography gently reduces moving gineering bedrock. However, assessing local seismic amplifications in
southwest (Fig. 5). Otherwise, the BR roof height gradually increases in the Castelnuovo area requires developing 2D or 3D numerical models
the same direction (CN1 and CN2 in Fig. 4). Both topography and BR that must be sufficiently constrained with geotechnical and geophysical
height contribute to the thinning of the SI lithology towards the south­ surveys, to provide a more detailed picture of the bedrock shape and the
west and the consequent increase of frequency observed in Fig. 6e. soil linear and nonlinear dynamic properties at a smaller scale with
This explanation is confirmed by the performed 1D linear dynamic respect to that provided in our study. Such an aspect is fundamental
analyses (Fig. 9). Indeed, the modelled stiffness contrast between the SI since basin geometry strongly affects the reflection of seismic waves and,
and BR lithology produces a 1D elastic transfer function whose 1st consequently, their amplitude and frequency content. However, our
resonant frequency values are compatible with those observed with analysis proved that the availability of in-situ geological and geotech­
microtremor surveys. nical data (e.g., deep boreholes and geophysical measurements) and its
Assuming a power-law relationship between the SI thickness and the joint interpretation is fundamental to assess the dynamic properties of
resonant frequency of the type h = af0 b , (Ibs-Von Seht and Wohlenberg, soils and the local seismic site amplification (Bajaj and Anbazhagan,
1999; D’Amico et al., 2008), where h (m) is the SI thickness at boreholes 2019; Chandran and Anbazhagan, 2017).
CN1 and CN2 (Fig. 9), and f0 is the resonant frequency at S1 and S39 Indeed, our 1st-order results provide an updated picture of the sub­
HVSR curves (Fig. 9), we can provide a rough estimate of a and b co­ soil beneath the Castelnuovo hill that substantially differs from previous
efficients equal to 185.22 m/s and − 1.192, respectively. Although such studies We show for the first time that the observed HVSR frequency
coefficients are calibrated with two couples of measurements only, their peaks are associated with an impedance contrast at considerable depth
values and the obtained curve fall within the ranges estimated with the (210 m) at the lithological passage between clayey silts (SI) and car­
same empirical relation in other quaternary basins in Italy and world­ bonate breccias (BR). Such an impendence contrast was barely taken to
wide (supplementary Fig. S4). This relationship provides a 1st-order account or considered shallower, i.e., less than 150 m below the Cas­
picture of the trends of SI thickness and the BR roof height along two telnuovo hill, in previous studies (Lanzo et al., 2011a; Sica et al., 2014;
cross-sections A–B and C–D in Fig. 10. Landolfi, 2013; Evangelista et al., 2016). Moreover, the CN1 and CN2
Moving from northwest to southeast (A-B cross-section in Fig. 10), boreholes (Fig. 4) and the interpolated resonant period map (Fig. 8a)
the resonant period (red line in Fig. 10a) shows an overall reduction show that the SI lithology tapers to the east-southeast direction to
from 1.46 s to about 0.34 s, with a mean rate of approximately − 8.03 × approximately 50 m (Fig. 10a), while the BR roof height increases in the
10− 4 s/m. The rate of period change is not constant (blue line in Fig. 10). same direction. In contrast, in previous studies, the SI thickness was
Still, it shows a deceleration approaching the northwestern side of the assumed increasing to more than 100 m to the southwest direction, with
Castelnuovo hill to approximately − 1.6 × 10− 4 s/m, and a successive the engineering bedrock height decreasing in the same direction (Lanzo
acceleration moving towards the southeastern side with a peak of − 1.5 et al., 2011a; Sica et al., 2014; Landolfi, 2013; Evangelista et al., 2016).
× 10− 3 s/m. The rate variation in the proximity of the hill is due to both Therefore, the performed analyses and the geo-lithological profiles in
the inclined roof of the BR engineering bedrock testified by the CN1 and Fig. 10 provide an updated basis for the implementation of 2D numerical
CN2 boreholes and to the topography changes, which make the SI li­ models and the calculation of Castelnuovo hill’s amplification effects,
thology ticker below the Castelnuovo hill, thus increasing the resonant which is mandatory for estimating the contribution of stratigraphic
period. topographic effects. Moreover, the SI lithology shows resonant fre­
In the SE-NW direction (C-D cross-section in Fig. 10), the resonant quencies ranging between 0.7 and 3.0 Hz. The latter represents the
period varies from 0.97 s on the southeast to 0.61 on the northwest (the typical resonant frequencies of ordinary buildings in civil engineering;
red curve in Fig. 10), with a mean rate of approximately − 3.1 × 10− 4 s/ therefore, possible double resonance effects could be expected during an
m. Here, the maximum resonant period (about 1.12 s) is attained along earthquake, thus greatly amplifying seismic waves (Tallini et al., 2020).
the northeastern side of the Castelnuovo hill and is shifted with respect
to the hilltop. Otherwise, the rate of period change (the blue curve in 6. Conclusions
Fig. 10) becomes positive in the proximity of the hilltop (7.5 × 10− 4 s/
m) and negative at the base of the hill to the northeast (− 1.3 × 10− 3 s/ This study provided for the first time the results of a joint analysis
m). Such variations reflect the changes in SI thickness due to both the between geophysical measurements and deep boreholes reaching the
topography changes in the proximity of the Castelnuovo hill and the engineering bedrock for the Castelnuovo area. The results, which are
possible presence of fault scarps or morphological lines that dislocate summarized in the two orthogonal cross-sections in Fig. 10, indicate that
and lower the BR in the northeastern part of the cross-section. resonance effects occur in the frequency range of 0.7–3.0 Hz due to the
As per the local seismic amplification of the study area, microtremor presence of a velocity contrast related to the stratigraphic boundary
measurements for f0 < 3.8 Hz show significant peak amplitudes up to 10, between the clayey silts of the San Nicandro Formation (SI) and the
related to the impedance contrast at depth, without an evident rela­ underlying the breccias (BR), whose depth and shape are different than
tionship between HVSR amplitude and altitude (supplementary Fig. S3). hypothesized up to now. Such an impedance contrast is deemed as the
Such a result suggests that stratigraphic amplification is more critical main factor for local amplification with respect to the topographic
than topographic amplification. However, more data are required to effect.
quantify both contributions to the local seismic amplification. This study could be denoted as a milestone for assessing the seismic

13
M. Spadi et al. Engineering Geology 297 (2022) 106506

hazard of the Castelnuovo village and demonstrates the need for basin (Central Italy). In: IAS fieldtrip guide, 34th IAS Meeting of Sedimentology
2019, Rome, Italy.
multidisciplinary in-situ geological investigations to study seismic site
D’Amico, V., Picozzi, M., Baliva, F., Albarello, D., 2008. Ambient noise measurements for
amplification. preliminary site-effects characterization in the Urban area of Florence. Italy. Bull.
Seismol. Soc. Am. 98, 1373–1388. https://doi.org/10.1785/0120070231.
Del Monaco, F., Tallini, M., De Rose, C., Durante, F., 2013. HVNSR survey in historical
Declaration of Competing Interest downtown L’Aquila (Central Italy): Site resonance properties vs. subsoil model. Eng.
Geol. 158, 34–47. https://doi.org/10.1016/j.enggeo.2013.03.008.
Di Naccio, D., Famiani, D., Liberi, F., Boncio, P., Cara, F., De Santis, A., Di Giulio, G.,
The authors declare that they have no known competing financial Galadini, F., Milana, G., Rosatelli, G., Vassallo, M., 2020. Site effects and widespread
interests or personal relationships that could have appeared to influence susceptibility to permanent coseismic deformation in the Avezzano town (Fucino
basin, Central Italy): Constraints from detailed geological study. Eng. Geol. 270,
the work reported in this paper. 105583 https://doi.org/10.1016/j.enggeo.2020.105583.
Evangelista, L., Landolfi, L., d’Onofrio, A., Silvestri, F., 2016. The influence of the 3D
morphology and cavity network on the seismic response of Castelnuovo hill to the
Acknowledgements
2009 Abruzzo earthquake. Bull. Earthq. Eng. 14, 3363–3387. https://doi.org/
10.1007/s10518-016-0011-8.
The deep borehole CN1 was funded by the following institutions, Galli, P.A.C., Giaccio, B., Messina, P., Peronace, E., Zuppi, G.M., 2011. Palaeoseismology
which are warmly thanked for their support: Dipartimento di Ingeg­ of the L’Aquila faults (Central Italy, 2009, Mw 6.3 earthquake): Implications for
active fault linkage. Geophys. J. Int. 187, 1119–1134. https://doi.org/10.1111/
neria, Edile-Architettura e Ambientale, Università dell’Aquila (Italy); j.1365-246X.2011.05233.x.
Dipartimento di Scienze, Università di Roma Tre (Italy); Istituto di Gallipoli, M.R., Albarello, D., Mucciarelli, M., Bianca, M., 2011. Ambient noise
Geologia Ambientale e Geoingegneria, CNR (Italy); Ufficio Speciale per measurements to support emergency seismic microzonation: the Abruzzo 2009
earthquake experience. Boll. Geofis. Teor. Appl. 52, 539–559. https://doi.org/
la Ricostruzione dei Comuni Cratere (USRC), Fossa (L’Aquila, Italy). 10.4430/bgta0031.
Giorgio Pipponzi and USRC are thanked for the local logistic assistance. Gallipoli, M.R., Bianca, M., Mucciarelli, M., Parolai, S., Picozzi, M., 2013. Topographic
We also would like to thank Luca Maria Puzzilli for the stimulating versus stratigraphic amplification: Mismatch between code provisions and
observations during the L’Aquila (Italy, 2009) sequence. Bull. Earthq. Eng. 11,
discussion on the down hole tests. 1325–1336. https://doi.org/10.1007/s10518-013-9446-3.
Gallipoli, M.R., Calamita, G., Tragni, N., Pisapia, D., Lupo, M., Mucciarelli, M., Stabile, T.
A., Perrone, A., Amato, L., Izzi, F., La Scaleia, G., Maio, D., Salvia, V., 2020.
Appendix A. Supplementary data
Evaluation of soil-building resonance effect in the urban area of the city of Matera
(Italy). Eng. Geol. 272, 105645 https://doi.org/10.1016/j.enggeo.2020.105645.
Supplementary data to this article can be found online at https://doi. Giaccio, B., Galli, P., Messina, P., Peronace, E., Scardia, G., Sottili, G., Sposato, A.,
org/10.1016/j.enggeo.2021.106506. Chiarini, E., Jicha, B., Silvestri, S., 2012. Fault and basin depocentre migration over
the last 2 Ma in the L’Aquila 2009 earthquake region, central Italian Apennines.
Quat. Sci. Rev. 56, 69–88. https://doi.org/10.1016/j.quascirev.2012.08.016.
References Gosar, A., Lenart, A., 2010. Mapping the thickness of sediments in the Ljubljana Moor
basin (Slovenia) using microtremors. Bull. Earthq. Eng. 8, 501–518. https://doi.org/
10.1007/s10518-009-9115-8.
Albarello, D., Castellaro, S., 2011. Tecniche sismiche passive: indagini a stazione singola.
Guidoboni, E., Comastri, A., Mariotti, D., Ciuccarelli, C., Bianchi, M.G., 2012. Ancient
Ing. sismica 38, 32–49.
and Medieval Earthquakes in the Area of L’Aquila (Northwestern Abruzzo, Central
Amanti, M., Muraro, C., Roma, M., Chiessi, V., Puzzilli, L.M., Catalano, S., Romagnoli, G.,
Italy), A.D. 1–1500: a critical revision of the Historical and Archaeological Data.
Tortorici, G., Cavuoto, G., Albarello, D., Fantozzi, P.L., Paolucci, E., Pieruccini, P.,
Bull. Seismol. Soc. Am. 102 (4), 1600–1617. https://doi.org/10.1785/0120110173.
Caprari, P., Mirabella, F., Della Seta, M., Esposito, C., Di Curzio, D., Francescone, M.,
Ibs-Von Seht, M., Wohlenberg, J., 1999. Microtremor Measurements used to Map
Pizzi, A., Macerola, L., Nocentini, M., Tallini, M., 2020. Geological and geotechnical
Thickness of Soft Sediments. Bull. Seismol. Soc. Am. 89, 250–259.
models definition for 3rd level seismic microzonation studies in Central Italy. Bull.
Ilic, V., Del, E., Ghinelli, A., Boschi, S., Durante, F., Pipponzi, G., Ilic, V., Del, E.,
Earthquake Eng. Springer Netherlands. https://doi.org/10.1007/s10518-020-
Ghinelli, A., Boschi, S., Durante, F., Pipponzi, G., 2017. Local Seismic Response of
00843-x.
Castelnuovo Hill (AQ-Italy) with 2D models, 56–67. ANIDIS 2017, Pistoia, Italy.
Ameri, G., Oth, A., Pilz, M., Bindi, D., Parolai, S., Luzi, L., Mucciarelli, M., Cultrera, G.,
Konno, K., Ohmachi, T., 1998. Ground-motion characteristics estimated from spectral
2011. Separation of source and site effects by generalized inversion technique using
ratio between horizontal and vertical components of microtremor. Bull. Seismol.
the aftershock recordings of the 2009 L’Aquila earthquake. Bull. Earthq. Eng. 9,
Soc. Am. 88, 228–241.
717–739. https://doi.org/10.1007/s10518-011-9248-4.
Kottke, A.R., Wang, X., Rathje, E.M., 2019. Strata Technical Manual.
Amoroso, S., Gaudiosi, I., Tallini, M., Di Giulio, G., Milana, G., 2018. 2D site response
Krige, D.G., 1951. A statistical approach to some basic mine valuation problems on the
analysis of a cultural heritage: the case study of the site of Santa Maria di
Witwatersrand. J. South. Afr. Inst. Min. Metall. 52, 119–139.
Collemaggio Basilica (L’Aquila, Italy). Bull. Earthq. Eng. 16, 4443–4466. https://doi.
Lai, C.G., Poggi, V., Famà, A., Zuccolo, E., Bozzoni, F., Meisina, C., Bonì, R., Martelli, L.,
org/10.1007/s10518-018-0356-2.
Massa, M., Mascandola, C., Petronio, L., Affatato, A., Baradello, L., Castaldini, D.,
Antonielli, B., Della Seta, M., Esposito, C., Scarascia Mugnozza, G., Schilirò, L., Spadi, M.,
Cosentini, R.M., 2020. An inter-disciplinary and multi-scale approach to assess the
Tallini, M., 2020. Quaternary rock avalanches in the Apennines: New data and
spatial variability of ground motion for seismic microzonation: the case study of
interpretation of the huge clastic deposit of the L’Aquila Basin (Central Italy).
Cavezzo municipality in Northern Italy. Eng. Geol. 274 https://doi.org/10.1016/j.
Geomorphology 361, 107194. https://doi.org/10.1016/j.geomorph.2020.107194.
enggeo.2020.105722.
Bajaj, K., Anbazhagan, P., 2019. Seismic site classification and correlation between V S
Landolfi, L., 2013. Analisi della risposta sismica locale in condizioni complesse di
and SPT-N for deep soil sites in Indo-Gangetic Basin. J. Appl. Geophys. 163, 55–72.
sottosuolo: il caso di Castelnuovo (AQ), PhD Thesis in Geotechnical Engineering,
https://doi.org/10.1016/j.jappgeo.2019.02.011.
Universita` di Napoli Federico II (in Italian).
Chandran, D., Anbazhagan, P., 2017. Subsurface profiling using integrated geophysical
Lanzo, G., Silvestri, F., Costanzo, A., d’Onofrio, A., Martelli, L., Pagliaroli, A., Sica, S.,
methods for 2D site response analysis in Bangalore city, India: a new approach.
Simonelli, A., 2011a. Site response studies and seismic microzoning in the Middle
J. Geophys. Eng. 14, 1300–1314. https://doi.org/10.1088/1742-2140/aa7bc4.
Aterno valley (L’aquila, Central Italy). Bull. Earthq. Eng. 9, 1417–1442. https://doi.
Chiarabba, C., Jovane, L., DiStefano, R., 2005. A new view of Italian seismicity using 20
org/10.1007/s10518-011-9278-y.
years of instrumental recordings. Tectonophysics 395, 251–268. https://doi.org/
Lanzo, G., Tallini, M., Milana, G., Di Capua, G., Del Monaco, F., Pagliaroli, A.,
10.1016/j.tecto.2004.09.013.
Peppoloni, S., 2011b. The Aterno valley strong-motion array: Seismic
Chiarabba, C., Amato, A., Anselmi, M., Baccheschi, P., Bianchi, I., Cattaneo, M.,
characterization and determination of subsoil model. Bull. Earthq. Eng. 9,
Cecere, G., Chiaraluce, L., Ciaccio, M.G., De Gori, P., De Luca, G., Di Bona, M., Di
1855–1875. https://doi.org/10.1007/s10518-011-9301-3.
Stefano, R., Faenza, L., Govoni, A., Improta, L., Lucente, F.P., Marchetti, A.,
Maresca, R., Berrino, G., 2016. Investigation of the buried structure of the Volturara
Margheriti, L., Mele, F., Michelini, A., Monachesi, G., Moretti, M., Pastori, M., Piana
Irpina Basin (southern Italy) by microtremor and gravimetric data. J. Appl. Geophys.
Agostinetti, N., Piccinini, D., Roselli, P., Seccia, D., Valoroso, L., 2009. The 2009
128, 96–109. https://doi.org/10.1016/j.jappgeo.2016.03.010.
L’Aquila (Central Italy) Mw6.3 earthquake: main shock and aftershocks. Geophys.
Martelli, L., Boncio, P., Baglione, M., Cavuoto, G., Mancini, M., Mugnozza, G.S.,
Res. Lett. 36, 1–6. https://doi.org/10.1029/2009GL039627.
Tallini, M., 2012. Main geologic factors controlling site response during the 2009
Cosentino, D., Cipollari, P., Marsili, P., Scrocca, D., 2010. Geology of the central
L’Aquila earthquake. Ital. J. Geosci. 131, 423–439. https://doi.org/10.3301/
Apennines: a regional review. J. Virtual Explor. 36 https://doi.org/10.3809/
IJG.2012.12.
jvirtex.2010.00223.
Meletti, C., Montaldo, V., 2007. Stime di pericolosità sismica per diverse probabilità di
Cosentino, D., Asti, R., Nocentini, M., Gliozzi, E., Kotsakis, T., Mattei, M., Esu, D.,
superamento in 50 anni: valori di ag. Progetto Dipartimento di Protezione Civile
Spadi, M., Tallini, M., Cifelli, F., Pennacchioni, M., Cavuoto, G., Di Fiore, V., 2017.
(DPC), Istituto Nazionale di Geofisica e Vulcanologia (INGV) S1, Deliverable D2,
New insights into the onset and evolution of the central Apennine extensional
Milano, Italy. http://esse1.mi.ingv.it/d2.html (last access April 2021).
intermontane basins based on the tectonically active L’Aquila Basin (Central Italy).
Molnar, S., Cassidy, J.F., Castellaro, S., Cornou, C., Crow, H., Hunter, J.A.,
Bull. Geol. Soc. Am. 129, 1314–1336. https://doi.org/10.1130/B31679.1.
Matsushima, S., Sánchez-Sesma, F.J., Yong, A., 2018. Application of microtremor
Cosentino, D., Giaccio, B., Gliozzi, E., Nocentini, M., Pipponzi, G., Spadi, M., Tallini, M.,
horizontal-to-vertical spectral ratio (MHVSR) analysis for site characterization: state
2019. Lacustrine deposits of the late Piacenzian-Gelasian L’Aquila intermontane

14
M. Spadi et al. Engineering Geology 297 (2022) 106506

of the art. Surv. Geophys. 39, 613–631. https://doi.org/10.1007/s10712-018-9464- SESAME, 2004. Guidelines for the implementation of the H/V spectral ratio technique on
4. ambient vibrations-measurements, processing and interpretations. In: SESAME
Nakamura, Y., 1989. A method for dynamic characteristics estimation of subsurface European research project EVG1-CT-2000-00026, deliverable D23.12. SESAME Site
using microtremor on the ground surface. Railw. Tech. Res. Institute, Q. Rep. 30. Eff. Assess. using Ambient Excit, 1–62.
Nakamura, Y., 2019. What is the Nakamura method? Seismol. Res. Lett. 90, 1437–1443. Sica, S., Dello Russo, A., Rotili, F., Simonelli, A.L., 2014. Ground motion amplification
Nocentini, M., Cosentino, D., Spadi, M., Tallini, M., 2018. Plio-quaternary geology of the due to shallow cavities in nonlinear soils. Nat. Hazards 71, 1913–1935. https://doi.
paganica-san demetrio-castelnuovo basin (Central Italy). J. Maps 14, 411–420. org/10.1007/s11069-013-0989-z.
https://doi.org/10.1080/17445647.2018.1481774. Spadi, M., Gliozzi, E., Cosentino, D., Nocentini, M., 2016. Late Piacenzian–Gelasian
Ohsaki, Y., Iwasaki, R., 1973. On dynamic shear moduli and poisson’s ratios of soil freshwater ostracods (Crustacea) from the L’Aquila Basin (central Apennines, Italy).
deposits. Soils Found. 13, 61–73. https://doi.org/10.3208/sandf1972.13.4_61. J. Syst. Palaeontol. 14, 617–642. https://doi.org/10.1080/
Oliver, M.A., Webster, R., 2014. A tutorial guide to geostatistics: Computing and 14772019.2015.1079561.
modelling variograms and kriging. Catena 113, 56–69. https://doi.org/10.1016/j. Storti, F., Aldega, L., Balsamo, F., Corrado, S., Del Monaco, F., Di Paolo, L., Mastalerz, M.,
catena.2013.09.006. Monaco, P., Tallini, M., 2013. Evidence for strong middle Pleistocene earthquakes in
Oubaiche, E.H., Chatelain, J.L., Hellel, M., Wathelet, M., MacHane, D., Bensalem, R., the epicentral area of the 6 April 2009 L’Aquila seismic event from sediment
Bouguern, A., 2016. The relationship between ambient vibration H/V and SH paleofluidization and overconsolidation. J. Geophys. Res. Solid Earth 118,
transfer function: some experimental results. Seismol. Res. Lett. 87, 1112–1119. 3767–3784. https://doi.org/10.1002/jgrb.50254.
https://doi.org/10.1785/0220160113. Tallini, M., Cavuoto, G., Del Monaco, F., Di Fiore, V., Mancini, M., Caielli, G.,
Panzera, F., Romagnoli, G., Tortorici, G., D’Amico, S., Rizza, M., Catalano, S., 2019. Cavinato, G.P., De Franco, R., Pelosi, N., Rapolla, A., 2012. Seismic surveys
Integrated use of ambient vibrations and geological methods for seismic integrated with geological data for in-depth investigation of Mt. Pettino active Fault
microzonation. J. Appl. Geophys. 170, 103820 https://doi.org/10.1016/j. area (Western L’Aquila Basin). Ital. J. Geosci. 131, 389–402. https://doi.org/
jappgeo.2019.103820. 10.3301/IJG.2012.10.
Parolai, S., 2012. Investigation of site response in urban areas by using earthquake data Tallini, M., Spadi, M., Cosentino, D., Nocentini, M., Cavuoto, G., Di Fiore, V., 2019. High-
and seismic noise. In: New Manual of Seismological Observatory Practice (NMSOP- resolution seismic reflection exploration for evaluating the seismic hazard in a Plio-
2), p. XX. https://doi.org/10.2312/GFZ.NMSOP-2. Quaternary intermontane basin (L’Aquila downtown, Central Italy). Quat. Int. 532,
Pebesma, E.J., 2004. Multivariable geostatistics in S: the gstat package. Comput. Geosci. 34–47. https://doi.org/10.1016/j.quaint.2019.09.016.
30, 683–691. Tallini, M., Sardo, L. Lo, Spadi, M., 2020. Seismic site characterisation of Red Soil and
Pitilakis, K., Raptakis, D., Lontzetidis, K., Tika-Vassilikou, T., Jongmans, D., 1999. soil-building resonance effects in L ’ Aquila downtown ( Central Italy). Bull. Eng.
Geotechnical and geophysical description of euro-seistest, using field, laboratory Geol. Environ. 79, 4021–4034. https://doi.org/10.1007/s10064-020-01795-x.
tests and moderate strong motion recordings. J. Earthq. Eng. 3, 381–409. https:// Toro, G.R., 1995. Probabilistic models of site velocity profiles for generic and site-specific
doi.org/10.1080/13632469909350352. ground-motion amplification studies. Tech. Rep. 779574.
QGIS Development Team, 2018. QGIS Geographic Information System. Open Source Toro, G.R., Silva, W.J., McGuire, R.K., Herrmann, R.B., 1992. Probabilistic seismic
Geospatial Foundation Project. hazard mapping of the Mississippi Embayment. Seismol. Res. Lett. 63, 449–475.
R Team Core, 2019. R: A Language and Environment for Statistical Computing (Version Trevisani, S., Boaga, J., Agostini, L., Galgaro, A., 2017. Insights into bedrock surface
3.0. 2)[Computer Software]. R Foundation for Statistical Computing, Vienna, morphology using low-cost passive seismic surveys and integrated geostatistical
Austria. analysis. Sci. Total Environ. 578, 186–202. https://doi.org/10.1016/j.
Rovida, A., Locati, M., Camassi, R., Lolli, B., Gasperini, P., Azzaro, R., Bernardini, F., scitotenv.2016.11.041.
Camassi, R., Amico, S.D., Ercolani, E., Locati, M., Rossi, A., Rovida, A., Trevisani, S., Pettenati, F., Paudyal, S., Sandron, D., 2021. Mapping long-period soil
Tertulliani, A., Gasperini, P., Lolli, B., Meletti, C., Rovida, A., Albini, P., Castelli, V., resonances in the Kathmandu basin using microtremors. Environ. Earth Sci. 80,
Caracciolo, C.H., Amico, V.D., Pondrelli, S., Rebez, A., Locati, M., Ingv, T., 2016. 1–16. https://doi.org/10.1007/s12665-021-09532-7.
Catalogo Parametrico dei Terremoti Italiani, versione CPTI15 1–33. Wathelet, M., Chatelain, J.L., Cornou, C., Di Giulio, G., Guillier, B., Ohrnberger, M.,
Saroli, M., Lancia, M., Modoni, G., Moro, M., Mugnozza, G.S., Lincei, R., 2014. New Savvaidis, A., 2020. Geopsy: a user-friendly open-source tool set for ambient
geological data on the Cassino intermontane basin, central Apennines, Italy. Rend. vibration processing. Seismol. Res. Lett. 91, 1878–1889. https://doi.org/10.1785/
Lincei 25, 189–196. https://doi.org/10.1007/s12210-014-0338-5. 0220190360.
Saroli, M., Albano, M., Modoni, G., Moro, M., Milana, G., Spacagna, R.L., Falcucci, E., Working Group MS-AQ, 2010. La Microzonazione sismica dell’area aquilana. Vol. 3 +
Gori, S., Scarascia Mugnozza, G., 2020. Insights into bedrock paleomorphology and dvd. Dipartimento della Protezione Civile-Regione Abruzzo (in Italian: Seismic
linear dynamic soil properties of the Cassino intermontane basin (Central Italy). Eng. microzonation for the building reconstruction of L’Aquila area. Abruzzo Region –
Geol. 264 https://doi.org/10.1016/j.enggeo.2019.105333. Italian Civil Protection Department).
Seed, H.B., Idriss, I.M., 1981. Evaluation of liquefaction potential sand deposits based on
observation of performance in previous earthquakes. ASCE National Convention
(MO) 81–544.

15

You might also like