GeolCarp_Vol56_No4_297

GEOLOGICA CARPATHICA, AUGUST 2005, 56, 4, 297306

www.geologicacarpathica.sk

Effects of lithological features and tectonic structures in the

evaluation of local seismic response: an example from the

Hyblean Plateau (Eastern Sicily)

ROSARIA RIGANO and GIUSEPPE LOMBARDO

Dipartimento di Scienze Geologiche, University of Catania, C.so Italia 55, 95129 Catania, Italy; riganor@unict.it; lombardo@unict.it

(Manuscript received July 1, 2004; accepted in revised form December 9, 2004)

Abstract: The influence of faults and different lithology in the evaluation of possible seismic energy amplification has

been studied. The area of Augusta, located in the Hyblean Plateau (south-eastern Sicily), was selected as test site and

several ambient noise measurements were carried out with the aim of investigating the local seismic response. Data were

gathered both on the main lithotypes outcropping in the study area and along some profiles crossing the Mt Tauro fault,

in order to estimate the site response connected to the different geo-lithological features and to investigate on the pos-

sible amplification effects caused by the presence of a tectonic structure. The horizontal to vertical (H/V) spectral ratios

calculated for the sites located near the fault showed a rough tendency towards amplifications in the frequency band 1.5

2.0 Hz, disappearing with increasing distances from the fault. Significant amplifications were observed at about 0.7 Hz

in the sites located on the clays and in all the lithotypes laying on them. Observed dominant spectral peaks have been

explained as the resonance of a soft sedimentary sequence overlaying an equivalent basement located inside the clayey

formation at a depth of 160290 m, or at the stratigraphic interface separating the clays from the calcareous basement.

Finally, the results of 1-D modelling allowed us to delimitate the trend of the Augusta graben at depth.

Key words: Hyblean Plateau, Sicily, Augusta, Mt Tauro fault, ambient noise, local seismic response, Nakamura technique.

Introduction

South-eastern Sicily is one of the Italian regions with the high-

est seismic hazard. It is characterized by high-energy releases

with events (M ≈ 67), which have caused heavy damage over

a territory, which especially in recent times was highly urban-

ized. The study area is located in the north-eastern boundary

of the Hyblean Plateau, which is a Mesozoic-Cenozoic car-

bonatic massif with repeated volcanic levels laying on the Af-

rican continental crust, as shown in the top right inset in

Fig. 1.

A series of horst and graben structures characterizes the tec-

tonic setting of this area. They are oriented NWSE and

NNWSSE and are linked to the Malta-Hyblean escarpment.

It is along this tectonic system that the strongest earthquakes

historically affecting south-eastern Sicily took place. The

1169 and 1693 destructive earthquakes belong indeed to the

coastal sector of Hyblean Plateau and originated in such re-

gional scale tectonic structures. According to recent papers

(Barbano & Rigano 2001; Barbano et al. 2001a) the town of

Augusta is one of the localities with the highest probability of

occurrence of destructive events. The instrumental data refer

to the 19861995 time interval (Salvi et al. 1996) and describe

the seismicity in terms of small magnitude events that, in most

of the cases, are located along the eastern coastal sector. The

most recent large earthquake (M = 5.4) was located in the Ion-

ian Sea at a depth of about 10 km (Amato et al. 1995). It oc-

curred on December 13

1990 and it is the only event for

which instrumental parameters are available (Giardini et al.

1995). In spite of the moderate magnitude of the 1990 event,

the town was seriously damaged and the VIII degree of the

MCS scale was here found.

In the study area it is possible to distinguish a bedrock and

an overlaying lithostratigraphic sequence, whose geological

characteristics and geometry can potentially induce variations

in the seismic energy recorded at different shallow sites. Sig-

nificant differences in the site response of different lithotypes

laying on the bedrock were preliminarily observed by Barbano

et al. (2001b).

It is a well-documented phenomenon that earthquake

ground motion can be amplified by local site conditions (Aki

1988). Near-surface impedance contrasts between bedrock

and shallower deposits can significantly affect the frequency-

amplitude content and the duration of earthquake ground mo-

tion. A case history illustrating such a phenomenon occurred

during the 1985 Michoacan, Mexico, earthquake (M = 8).

This event, despite its significant epicentral distance, excited

the ~2 sec fundamental vibration mode of the Mexico City

lake bed zone, causing severe damage and thousands of deaths

(Anderson et al. 1986; Singh et al. 1989). Recently, strong ev-

idence that site response contributed to enhancing the damage

was observed in the town of Nocera Umbra, during the 1997

Umbria-Marche earthquake (Donati et al. 2001). The highest

damage grades were estimated for two zones, on the soft sedi-

ment of a river valley and on the terrains within a 200 m wide

fault zone crossing the town.

The aim of this paper is to determine the seismic site re-

sponse of the Augusta urbanized area, by investigating both

the local response of sites with different soil conditions and in

the neighbourhoods of a tectonic structure. The focusing of

298 RIGANO and LOMBARDO

Fig. 1. Geolithological map of the Augusta area. The upper right inset map shows a tectonic sketch of eastern Sicily.

seismic waves by tectonic discontinuities can indeed cause

spatial variations of ground motion in both amplitude and fre-

quency since stationary waves are generated by the impedance

contrast between the cataclastic area and the surroundings

rocks (Li & Leary 1990).

The methodology adopted involves the use of ambient noise

and the comparison of the spectral ratios with the results from

a 1-D modelling in order to evaluate the resonant frequency of

the soft cover layers.

Geological setting

Augusta urban area is located in the Hyblean Plateau, which

represents the emerged foreland elevated with respect to its

surroundings. It partially stands on a little island where ter-

rains of lower Pleistocene age, prevailingly formed by cal-

carenites and clays, outcrop. The whole lithostratigraphic se-

quence which characterizes this area consists of alternating

sedimentary deposits and pyroclastic layers laying on a Mi-

ocene carbonatic formation well known as the Syracuse Lime-

stone Member of the Mt Climiti Formation (Pedley 1981).

This last overlies unconformably on Cretaceous lavas mostly

represented by pyroclastics, and sometimes by highly altered

lavas. The geolithological features of the study area are shown

in the map of Fig. 1, redrawn from Lentini et al. (1986).

The distribution and the geometric features of the faults in

the north-eastern sector of the Hyblean Plateau are connected

with the lower Pleistocene alternating tectonic movements of

uplift and lowering, which imply a structural setting with horst

and graben. In this framework, the NNWSSE trending fault

structures of Mt Tauro horst are arranged in a typical en-eche-

lon pattern. They are linked, together with the Augusta gra-

ben, to the Malta Escarpment strike-slip component of motion

(Grasso 1993).

The geological features of the investigated area were also

identified by collecting all available information from existing

boreholes. Such data were used to draw the geological cross-

section shown in Fig. 2. It depicts the average thickness of the

main lithotypes outcropping in the study area. Down-hole and

SASW (spectral analysis of surface waves) measurements

were used as well in order to characterize the geophysical and

geotechnical properties of various lithotypes. In addition to

these data an electrical tomography profile (160 m) was per-

LITHOLOGICAL STRUCTURES IN THE EVALUATION OF SEISMIC RESPONSE (SICILY) 299

Fig. 2. Geological cross-section and results of electrical tomography in the Mt Tauro fault area.

formed in the area where the Cretaceous pyroclastics outcrop

in order to investigate the features at the depth of the Mt Tauro

fault. The results (see inset profile in Fig. 2) show that at about

20 meters depth there is a lateral electric discontinuity that

marks a sharp change of the resistivity values from 108 to

17.3 ohm·m. The sudden lateral discontinuity between clays

and limestones can be explained by postulating a step down

fault, which lowers the clayey formation at the same depth of

the carbonatic and volcanic pyroclastic levels.

Methodology, data acquisition and processing

Microzoning of an earthquake prone area, such as the study

one, can be tackled by analysing the available information

about tectonic, geological and geotechnical features of all ex-

isting formations and lithotypes. The best procedure for deter-

mining the site response is to observe the ground motion dur-

ing an actual event. This can be done, using strong or weak

motion records, by direct comparison of a sediment site to a

reference one located on competent ground. Another alterna-

tive approach, first introduced by Kanai (1957), involves the

use of microtremors or ambient seismic noise. Among the

most used methods, the spectral ratio technique is very popu-

lar using either a reference rock station or taking the horizon-

tal-to-vertical component ratio. In recent times many authors

have demonstrated that the horizontal to vertical (H/V) spectral

ratios from ambient noise measurements are consistent in shape

with the ones coming from earthquake recordings and with

those obtained by using a reference bedrock station (Bard

1999). Therefore the use of the H/V spectral ratio of ambient

noise (Nakamura 1989) is widely adopted to characterize earth-

quake site response. This technique has been successfully used

to estimate the natural frequency of resonance of sedimentary

layers and recently extended by Lermo & Chávez-García

(1993) to evaluate sediment-induced amplification of S waves.

The best performance of the method is achieved when there is a

strong velocity contrast between the bedrock and the soft upper

layers. Moreover, the features of the layering of sediments itself

strongly affect the frequency band and the shape of the domi-

nant peaks in the H/V spectra (Fäh et al. 2001).

Several seismological studies did provide evidence of non-

linear behaviour of soft soils. Non-linear effects have the ten-

dency to reduce the amplitudes of the spectral peaks and shift

them towards slightly longer periods. Because of possible

non-linearities, the choice of weak-motion signals to be used

in predicting strong-motion behaviour remains a problem, and

the answer may vary according to soil type. However, as dis-

cussed by Aki (1988), recent studies show good correlation

between weak- and strong-motion site response.

We calculated the H/V spectral ratios of ambient noise re-

corded at different sites in the area of Augusta in order to esti-

mate the resonance frequency of sedimentary layers and to in-

vestigate the local seismic response due to the presence of a

tectonic structure.

A series of three component microtremor measurements

were sampled in 51 sites with different lithology (Fig. 1).

Measurements were also performed along five profiles cross-

ing the Mt Tauro fault. A total number of 29 additional sam-

pling sites was therefore selected for each profile, starting

from the fault scarp and moving apart from both edges of the

structure with steps of about 50 and 100 meters. The sites lo-

cated along the profiles are marked with both number and star

in order to distinguish them from the numbering of the sites

located on different lithotypes (Fig. 4).

Seismic signals were recorded using a 1-Hz Mark L4C 3-D

seismometer connected to a 12 bit analog-to-digital converter

and a notebook. Horizontal components of the geophone were

oriented parallel to the NS and EW geographic directions.

Sampling frequency was 100 Hz and two antialiasing filters

cut higher frequencies with a 10 db/oct slope. At each site five

Fig. 3. Spectral ratio obtained by averaging the H/V ratios over 5

time histories of 120 s recorded on the calcareous basement. The

mean spectral ratio is the dark curve; the upper and lower thin

curves represent the ±1 standard deviation interval.

LITHOLOGICAL STRUCTURES IN THE EVALUATION OF SEISMIC RESPONSE (SICILY) 301

time series of 120-s length were recorded. The time series

were baseline corrected in order to remove spurious offsets

and low-frequency trends. After the application of a Hanning

window a Fast Fourier Transform algorithm was applied, to

obtain spectra in a frequency band 0.120 Hz. The lower limit

of 0.1 Hz was assessed through the low-frequency cut off

where the microtremor standard deviation becomes equal to

the natural noise fluctuation of the 12-bit digitizer.

For each time history the vector composition of horizontal-

components noise spectra were divided by the vertical one. Fi-

nally the arithmetic average operation provides the mean H/V

spectral ratio at each measurement site. A running smoothing

function of 0.1 Hz was also used, both on the spectra of each

single component and on the final ratio, to reduce spectral

fluctuations.

Results

Before describing the results it is worth specifying that am-

bient noise measurements were performed in some sites locat-

ed on the bedrock (limestones) in order to obtain a univocal

standard for defining the significant amplification spectral

peaks. The standard deviation evaluated for the H/V produces

spectral fluctuations of ±0.5 units, around mean amplitude

values not exceeding 3.0 units (Fig. 3). On the basis of these

Fig. 5. H/V spectral ratios computed from measurements performed in sites located on the limestones (a) and on the calcarenitic terraces

(Ms, Qs and Mvc) overlaying them (b).

fluctuations only spectral peaks higher than 3.5 units were tak-

en into account as statistically significant.

The schematic cross-sections shown in Fig. 4 refer to the

short profiles crossing the Mt Tauro fault. In each profile, H/V

spectral ratios obtained from noise measurements performed

in sites located at increasing distance from both edges of the

fault, are shown. It is possible to observe, in the frequency in-

terval 1.52.0 Hz, a tendency towards the amplification of the

spectral peaks in the sites located in proximity of the tectonic

structure. Although this is not a systematic behaviour, it is

more evident in the BB, CC, DD and EE profiles,

where the afore-mentioned spectral peaks tends to decrease

their amplitude at distance of about 3050 m from the fault.

Ambient noise was also sampled in 51 sites located on dif-

ferent soils outcropping in the whole territory of Augusta

(Fig. 1). The resulting H/V spectral ratios are grouped for sim-

ilar lithotypes and plotted in the Figs. 5, 6 and 7. The H/V at

sites located both on the limestones (Mc) (Fig. 5a) and on the

calcarenites (Ms, Qs, Mvc) (Fig. 5b) overlaying them show a

flat shape in the frequency range 0.515 Hz. Only at the sites

14, 15 and 17, amplification peaks at 1.52.0 Hz and 4.0

5.0 Hz, are more or less evident.

As regards the sampling sites located on the clays the H/V

spectral ratios show amplification at frequency values of about

0.7 and 1.5 Hz (Fig. 6a). Similar amplification peaks are ob-

served in almost all the spectral ratios obtained from data re-

302 RIGANO and LOMBARDO

corded in the lithotypes laying on the clays, such as the cal-

carenites Qm (Fig. 6b) and salt pits (Fig. 7a). These terrains

have a modest (not exceeding ten meters) thickness compared

to that of the clayey formation.

The H/V obtained from ambient noise measurements in

sites located on the alluvial deposits show amplification peaks

at values ranging between 1.5 and 2.8 Hz (Fig. 7b).

1-D modelling and evaluation of soft sediments thickness in

the Augusta graben

In order to quantify the effect of the local stratigraphy, spec-

tral peaks experimentally observed were compared with the

Fig. 6. H/V spectral ratios computed from measurements performed in sites located on the clays (a) and on the calcarenites (Qm) over-

laying them (b).

results of a 1-D modelling, using the Dobry et al. (1976) and

the Haskell-Thomson methods. The method of Dobry, well

known as the Successive use of two layers method was used

both to test the reliability of the obtained results and to inter-

pret them from a geological-structural point of view. The

physical model adopted is a resonant cavity closed at its lower

extremity, which generally coincides with the contact between

the basement and the overlying formations. The basement

has, in this case, a physical meaning rather than the classical

geological one. It refers indeed to the physical interface under-

neath which the seismic input can be considered constant.

This is, in other words, the so called equivalent basement

(SEAOC 1974) indicating the interface, existing in any mate-

LITHOLOGICAL STRUCTURES IN THE EVALUATION OF SEISMIC RESPONSE (SICILY) 303

Fig. 7. H/V spectral ratios computed from measurements performed in sites located on the salt-pits (a) and on the alluvial deposits (b).

The fairly good relationship observed between the comput-

ed theoretical transfer function and the experimental dominant

H/V spectral peaks suggested that the authors should attempt a

correlation of the resonant frequencies experimentally detect-

Fig. 8. Theoretical transfer functions (dark lines) and H/V spectral

ratios (thin lines) obtained from the sites 48, 20 and 19 located on

clays, calcarenites and salt pits, respectively.

Table 1: Input data and results of 1-D modelling for three sites lo-

cated on clays, calcarenites and salt pits.

rial, underneath which the shear wave velocity at low defor-

mation (0.0001 %) reaches the value of V

=750 m/s.

Using this method it is possible to calculate the fundamental

period of a stratigraphic sequence by dividing it into a series

of layers pair. The fundamental period T of a layers pair is ob-

tained solving the equation:

where and are the fundamental

periods of the layers a and b, Ha, Hb, Va, Vb,

ρρρρρa and ρρρρρb are

the thicknesses, the shear wave velocities and the densities of

the layers a and b respectively. A detailed description and

some examples concerning the use of this method are given in

Lombardo (1987).

Through the Haskell-Thomson method the theoretical trans-

fer function for SH vertically incident waves was also comput-

ed, using as input data the thickness of the layers, the density

values and the velocities of transversal waves taken from

down-hole tests and SASW measurements available for the

study area. The input data and the results obtained through the

1-D modelling are listed in Table 1. In Fig. 8 the comparison

between the theoretical transfer function and three experimen-

tal H/V spectral ratios, is shown. These last were selected as

examples since they refer to the sampling sites closer to the lo-

cation of down-hole (see Fig. 1).

The theoretical site response obtained by using this model-

ling, shows a tendency to underestimate real data amplifica-

tions, but it gives resonant frequencies lower than 1 Hz, as ex-

perimentally observed in the H/V spectra. The 1-D modelling

shows indeed that dominant spectral peaks can be explained as

the resonance of a soft layer, having thickness ranging between

160 and 290 m, overlying an equivalent basement located ei-

ther inside the clayey formation or at the stratigraphic interface

separating clays from the calcareous basement (Table 1).

ñaHaTb

ñbHbTa

tan

4Ha

Ta =

4Hb

Tb =

304 RIGANO and LOMBARDO

In Fig. 9 a map obtained by plotting the values of thickness

of the sedimentary sequence filling the Augusta graben, esti-

mated through this equation, is shown. The map shows that

the clayey formation, widely outcropping in the Augusta gra-

ben, has a thickness ranging from 100 up to 300 meters going

from the edges to the central part of the basin.

Discussion and conclusions

The distribution of seismic energy depends on the complex-

ity of morphological and structural features of the investigated

area, as well as on its lithology. The different distribution of

earthquake damage in urban areas and, in general, of seismic

energy, depends indeed on the complexity of topographic,

structural and lithological features of the study areas. The fo-

cusing of seismic waves by tectonic discontinuities and the

trapping of the energy in basins can also cause spatial varia-

Fig. 9. Thickness contour map of the clayey formation in the Augusta graben.

ed with the thickness of the soft cover layers in the Augusta

graben. Such correlation is founded on the assumption that the

transfer function of a sediment site is connected to the geome-

try and physical parameters of the subsurface layers, namely,

their thickness and shear wave velocities (Ibs-von Seht &

Wohlenberg 1999). Theoretically the thickness (h) of a sedi-

mentary basin could be estimated from its resonant frequency

) and the shear wave velocity (V

) through the well known

relationship h = V

/4 f

. According to Ibs-von Seht & Wohlen-

berg (1999) in order to take into account the gradient of ve-

locity with depth, the relationship can be used,

where m is the mean thickness and f

(H/V)

is the main peak in

the Nakamura spectral ratio. The above relationship was tested

in the Lower Rhine Embayment. We adopted it in the study

area since the geotechnical parameters of Rhine Embayment

clays (V

= 303 m/s, ρ = 2.0 g/cm

) appear very similar to

those of the clayey sediments filling up the Augusta Basin

= 295 m/s, ρ = 1.9 g/cm

1.388

(H/V)

LITHOLOGICAL STRUCTURES IN THE EVALUATION OF SEISMIC RESPONSE (SICILY) 305

tions of ground motion in both amplitude and frequency. The

aim of this research was therefore to investigate the influence

of tectonic structures on the local seismic response and to de-

termine the resonant frequency of the soft cover layers.

Spectral ratios H/V have been computed for several sites in

the area of Augusta, located both on different soils and along

short profiles crossing the Mt Tauro fault. A tendency towards

site response amplification in the frequency range 1.52.0 Hz

was observed at the sites located close by the fault. Although

this is not a systematic behaviour, it clearly appears in some

profiles and such amplification tends to rapidly decrease for

distances ranging from about 30 to 50 meters from the struc-

tural discontinuity. Our methodological approach does not al-

low us to investigate the possible directional effects linked to

the strike of the fault. We rather directed our attention to rela-

tive differences between the amplitude of spectral peaks from

measurements close by the fault and at increasing distances

from it. Moreover, in order to simplify our observations we

deployed the profiles trying to locate them in portions of the

fault where it dislocates the same lithotype. The observed ten-

dency towards amplification effects in proximity of the Mt

Tauro fault appears therefore to be a significant consequence

of the impedance contrast between the gouge zone and the sur-

rounding rocks.

As regards the H/V obtained for the main lithotypes out-

cropping in the Augusta territory, at sites located both on the

limestones and on the calcarenites overlying them, a flat shape

of the spectral ratios in the frequency range 0.515 Hz, is ob-

served, therefore confirming the lack of amplification effects

in the lithotypes forming the basement. At sites located on the

clayey formation and on the thin sedimentary layers such as

salt-pits and quaternary calcarenites laying on them, signifi-

cant amplifications were observed at frequency values (0.7

1.5 Hz) that are compatible with the considerable thickness of

the clayey formation itself.

Comparison of experimental H/V dominant spectral peaks

with those obtained through the evaluation of the theoretical

transfer function, confirm that the resonance effect appears to

be related to the considerable thickness of the clayey forma-

tion, rather than to the thin shallow sediments. The equiva-

lent basement, detected through the 1-D modelling bounds

downwards the resonant soft sequence at a depth ranging from

160 up to 290 meters and it is located either among the clays

or at the stratigraphic interface separating them from the lime-

stones forming the basement.

The evaluation of the thickness of soft sediments infilling

the Augusta graben, shows that the maximum values of about

300 meters are reached in the central part of the basin. This re-

sult is in fairly good agreement with the thickness contour

map of the Pleistocene clayey formation, obtained from geo-

logical data by Carbone (1985).

Acknowledgment: This research was financially supported

by the Presidenza del Consiglio dei Ministri Dept. of Civil

Defence and the Presidenza della Regione Siciliana Sistema

Poseidon. The authors wish to thank Prof. A. Di Grande for

his support in describing and understanding the geological

features of the area. Special thanks are given to Prof. S. Gres-

ta for a series of useful discussions and suggestions both in the

planning and carrying out of the study. We are also grateful to

Prof. Riuscetti and to an anonymous reviewer for critical read-

ing of the manuscript and useful suggestions.

References

Aki K. 1988: Local site effects on strong ground motion. Proc.

Earthq. Engineering Soil Dynamics II 103155.

Amato A., Azzara R., Basili A., Chiarabba C., Cocco M., Di Bona

M. & Selvaggi G. 1995: Main shock and aftershocks of the

December 13, 1990, Eastern Sicily earthquake. Ann. di Geofi-

sica 38, 2, 255266.

Anderson J.G., Bodin P., Brune J.N., Prince J., Singh S.K., Quass

R. & Onate M. 1986: Strong ground motion from the Michoa-

can, Mexico earthquake. Science 233, 10431049.

Barbano M.S. & Rigano R. 2001: Earthquake sources and seismic

hazard in South-eastern Sicily. Ann. di Geofisica 44, 4, 723738.

Barbano M.S., Rigano R., Cosentino M. & Lombardo G. 2001a:

Seismic history and hazard in some localities of south-eastern

Sicily. Boll. Geof. Teor. Appl. 42, 12, 107120.

Barbano M.S., Di Grande A., Lombardo G. & Rigano R. 2001b:

Seismic hazard and site response in the area of Augusta. 20°

Convegno Nazionale GNGTS, Roma, poster section, 279281

(in Italian).

Bard P.Y. 1999: Microtremor measurements: a tool for site effects

estimation? In: Irikura K., Kudo K., Okada H. & Sasatani T.

(Eds.): The effects of Surface Geology on Seismic Motion.

Balkema, Rotterdam, 12511279.

Carbone S. 1985: The Pleistocene deposits of the north-eastern

Hyblean sector between Agnone and Melilli (SE Sicily): rela-

tionship between facies and structural elements. Boll. Soc.

Geol. Ital. 104, 405420 (in Italian).

Dobry R., Oweis I. & Urzua A. 1976: Simplified procedures for es-

timating the fundamental period of a soil profile. Bull. Seism.

Soc. Amer. 66, 4, 12931321.

Donati S., Marra F. & Rovelli A. 2001: Damage and ground shak-

ing in the town of Nocera Umbra during Umbria-Marche, cen-

tral Italy, earthquakes: the special effect of a fault zone. Bull.

Seism. Soc. Amer. 91, 3, 511519.

Fäh D., Kind F. & Giardini D. 2001: A theoretical investigation on

H/V ratios. Geophys. J. Int. 145, 535549.

Giardini D., Palombo B. & Pino N.A. 1995: Long-period model-

ling of MEDNET waveforms for the December 13, 1990 East-

ern Sicily earthquake. Ann. di Geofisica 38, 2, 267282.

Grasso M. 1993: Pleistocene structures along the Ionian side of the

Hyblean Plateau (SE Sicily): implications for the tectonic

evolution of the Malta Escarpment. In: Max M.D. & Colanto-

ni P. (Eds.): Geological development of the Sicilian-Tunisian

Platform. UNESCO Reports in Mar. Sci. 58, 4954.

Ibs-Von Seth M. & Wohlenberg J. 1999: Microtremor measure-

ments used to map thickness of soft sediments. Bull. Seism.

Soc. Amer. 89, 250259.

Kanai K. 1957: The requisite conditions for predominant vibration

of ground. Bull. Earthq. Res. Inst., Tokyo Univ. 31, 457.

Lentini F., Carbone S., Cugno G., Grasso M., Scamarda G., Sciuto

F., Montanari L., Romeo M. & Ferrara V. 1986: Geological

map of the north-eastern Hyblean sector. Ed. S.E.L.C.A. Map

scale 1:50,000, Firenze (in Italian).

Lermo J. & Chavez-Garcia F.J. 1994: Are microtremors useful in

site response evaluation? BSSA 84, 5, 13501364.

Li Y.G. & Leary P.C. 1990: Fault zone trapped seismic waves.