INSTITUTE OF PHYSICS PUBLISHING JOURNAL OF GEOPHYSICS AND ENGINEERING
J. Geophys. Eng. 3 (2006) 207–211 doi:10.1088/1742-2132/3/3/001
Seismic site effect estimation in the city of
Rabat (Morocco)
Saı̈d Badrane1, Lahcen Bahi2, Nacer Jabour1 and Aomar Iben Brahim1
1
Geophysics Laboratory, CNRST, BP 8027, Av. Omar Ibn al Khattab, Agdal, Rabat, Morocco
2
Ecole Mohammedia d’Ingénieurs, BP 765, Av. Ibn Sina Agdal Rabat, Morocco
Received 25 January 2006
Accepted for publication 22 May 2006
Published 26 June 2006
Downloaded from https://academic.oup.com/jge/article/3/3/207/5127598 by guest on 18 April 2025
Online at stacks.iop.org/JGE/3/207
Abstract
This study is based on a site survey carried out in the capital city of Morocco, Rabat. For this
aim, we used a technique based on the recording and processing of seismic ambient noise to
characterize local geological conditions in terms of the dynamic response of soil during
earthquakes. This technique consists of an assessment of the spectral ratio of the horizontal to
vertical components of microtremors recorded at the surface during a few minutes, and allows
determination of the fundamental period and the maximum amplification factor at the site of
measurement. We applied this technique in the city of Rabat to obtain distribution maps of
dominant periods and amplification factors assessed in more than 250 sites. In order to discuss
the results of this study, we used the information available in this zone about the lithology and
topography to correlate the obtained values of dominant periods and amplification factors with
the ground conditions. Finally, we established a microzoning map of the city of Rabat based
on the contours of dominant periods. The microzoning map established in this study can be
used by engineers and decision makers for urban and land use planning and also as a guide in
reduction of the seismic vulnerability of buildings.
Keywords: microtremors, dominant period, amplification
1. Introduction Several empirical and theoretical techniques (e.g. Field
and Jacob (1993), Lachet and Bard (1994), Duval et al (1996),
It is well known that the distribution of damage during Chavez-Garcia (1998), Bard (1998) and Nakamura (2000))
strong earthquakes is largely controlled by the site conditions. are used for the estimation of site response. The theoretical
The seismic waves radiated from the source are modified methods require detailed geotechnical information about the
by the upper soil layers before they reach the surface, materials of the upper soil layers which is not easy to obtain,
where structures and lifeline systems should withstand the especially in urban zones. The empirical techniques can be
amplitude and duration of the shaking caused by these seismic divided into two categories, the first includes techniques based
waves. on the recording of strong ground motion, and the second,
usually used in regions where the seismic activity is relatively
Microzoning based on identification of local
low, consists of techniques which use microtremor recording
modifications to ground motions due to the site conditions for assessing the transfer function of the soil.
represents the first step in any scheme for mitigating future In order to estimate the site effect in the city of Rabat,
losses from earthquakes in urban zones. The effect of which is characterized by low seismic activity, we used a
upper layers of soil on the seismic signal or the site effect simple and low cost method described by Nakamura (1989,
is expressed by two parameters, which are the dominant 2000). This method consists of recording urban seismic noise
period of the soil and the amplification factor. Knowledge of over a few minutes and dividing the spectrum of the horizontal
these parameters allows us to avoid the resonance of component by that of the vertical component of the ambient
structures and soil layers during earthquakes, and provides noise. This spectral ratio represents a new transfer function of
useful information for the design of earthquake resistant the site and allows us to determine the dominant period and
structures. the relative amplification value which corresponds to the peak
1742-2132/06/030207+05$30.00 © 2006 Nanjing Institute of Geophysical Prospecting Printed in the UK 207
S Badrane et al
Downloaded from https://academic.oup.com/jge/article/3/3/207/5127598 by guest on 18 April 2025
Figure 1. Location of the investigated area and its structural setting (Andrien 1974).
of the spectral ratio curve. Nakamura’s technique was tested Atlantic Ocean (Laouina et al 1994). The soil of Rabat is also
both experimentally and theoretically at many sites all over the characterized by the existence of some pockets filled by very
world by many authors (Duval et al 1996, Lachet et al 1994, soft sediments located with abundance in the sandstone.
Talhaoui et al 2004).
Following the same procedure, we used a database of noise
3. Data acquisition and processing
recorded in more than 250 sites in the urban zone of Rabat to
establish two maps that reflected the distribution of dominant We planned the site measurement grid of 500 × 500 m mesh on
periods and maximum relative amplifications. Based on the a detailed map of Rabat with a scale of 1/5000. Ambient noise
topographic map of Rabat, we assessed the altitude of each site was measured over 3 min in each site by means of a portable
measurement. Then we used the available information about seismograph Marslite 88 Lennartz technology, equipped with
the geology and topography to interpret the values of dominant a three component seismometer with a natural frequency of
periods and amplification factors obtained in this study. 1 Hz. Digital recordings were made with a sampling rate of
62.5 sps.
2. Geomorphologic and geological setting of In order to process the noise recorded at each site, as
Rabat described in the Nakamura papers (1989, 2000), we used
the program called SEISPECT written by Perelman and
The investigated area is situated along the Atlantic coastal plain Zaslavsky (2001), designed for the routine analysis of digital
(figure 1). It extends from the Bou Regreg River in the north to seismograms from three components seismic stations and
the province of Temara city in the south. This zone has been the digital accelerograms. This program with interactive windows
subject of many geomorphologic and geological studies (Akil enables us to view and determine the spectral ratio of horizontal
1980, Laouina et al 1994). In the middle Pliocene (Wernli components to that of vertical components in the recorded
1977), the Atlantic marine transgression moved forward along signal.
the coastal band of the meseta domain, depositing gravel The routine analysis used for assessing the dominant
and shelly sand which were transformed by diagenesis to period and its relative amplification in each site measurement
calcarenites and lenses of conglomerate. These sedimentary was as follows:
deposits slope toward the Atlantic Ocean and reach an altitude
of 250 m in the locality of Aı̈n el Aouda (figure 1). In the 1. Conversion of the N–S component, the E–W component
geomorphologic context, the surface topography of the studied and the vertical component of the recorded ambient noise
area is formed by the sandstone dunes and has an undulating to an appropriate format that can be analysed by the
shape in the direction of the Atlantic Ocean. SEISPECT program.
The studied area consists primarily of very hard rocks at 2. Choosing at least four windows of 20 s length on the three
the base (shale, crystalline rock and chalk), over which is set components.
a quaternary deposit of permeable limestone. At the surface 3. Calculation by SEISPECT, for each time window k, of the
we find sandstone and sandy soil called the red sand of Rabat. FFT of the horizontal components: (HN–S)k, (HE–W)k and
The thickness of the red soil decreases southwards toward the the vertical component: Vk .
208
Seismic site effect estimation in the city of Rabat
Downloaded from https://academic.oup.com/jge/article/3/3/207/5127598 by guest on 18 April 2025
Figure 2. Distribution of dominant periods overlaid on a geotechnical map of Rabat.
4. Assessment for each window, k, the spectral ratios of the (Andrien 1974). These values are represented in seven ranges,
two horizontal components (HN–S)k and (HE–W)k to the the highest values ranging from 1.4 to 1.5 s and the lowest
vertical component Vk . from 0.3 s to 0.4 s (figure 2).
5. Computation of the arithmetic average curve of all spectral The first observation was that the highest values were
ratios obtained in step 4. concentrated in the northern part of the investigated area,
n parallel to the Bou Regreg River. The values of dominant
1 (HN–S )k (HE–W )k
A= + periods are decreasing in a southerly direction. This seems in
2n k=1 Vk Vk agreement with the thickness variation of the upper soil layer
called red soil in the studied area. In fact, the thickness of this
where A = average spectral ratios curve, n = number of time
windows used in the computation, (HN–S)k and (HE–W)k = red soil is decreasing southwards towards the Atlantic Ocean,
spectra of the time window k in the horizontal components, as decribed in the geological and geomorphological setting of
Vk = spectrum of the time window k in the vertical component. Rabat (Laouina et al 1994).
This average spectral ratio curve obtained for each site of In some particular cases, the pick of the spectral ratio is
measurement, represents the approximate transfer function of not well exhibited; this phenomenon was observed at sites 6,
soil and allows us to assess the resonance frequency and its 10, 11 and 74, indicated by stars in figure 2. The seismic noise
relative amplification at the considered site. measurement at these sites may have been contaminated by
the effect of the intense urban traffic in this region.
We chose several sites of measurement in the investigated
4. Results and interpretation area in order to describe the spatial variation of the
Following the routine analysis of microtremors, as described assessed dominant periods and their correlation with the sites’
above, we assessed the resonance frequency and its relative geological conditions. The highest values of dominant periods
amplification factor in more than 250 sites over the urban were assessed in sites located in the area parallel to Bou Regreg
zone of Rabat. The obtained values of dominant periods are River (figure 2). This area is characterized by high altitude
between 0.3 s and 1.5 s, whereas the amplification values are and very thick sediment deposits. For illustration, we present
ranged from 1 to 12. in table 1 the altitudes, the geographical coordinates and the
obtained values of dominant periods for sites 1 and 5, which
are located in this zone. These high values for the dominant
4.1. Distribution of dominant periods over a geotechnical
periods assessed in these sites can be explained by the sites’
map of Rabat
location at high altitude in a region characterized by soft, thick
In order to correlate the values of dominant periods with the sediment deposits.
surface geology in the urban zone of Rabat, we plotted all the In the central part of the investigated area, the values of
dominant periods assessed on a geotechnical map of Rabat, dominant periods are around the value 0.7 s. We present in
209
S Badrane et al
Table 1. The values of dominant periods assessed in some sites.
Site Latitude Longitude Elevation Frequency Period
no (km) (km) (m) (Hz) (s) Amplification
1 382.39 366.98 33 0.7 1.4 2.0
5 382.06 367.05 35 0.7 1.4 2.4
112 378 365.93 50 1.4 0.7 3.4
53 377.50 365.50 46.5 1.4 0.7 3.5
150 377 371 8 2.5 0.4 4.1
155 377 371.5 6 2.8 0.3 3.4
331 376. 371 101 0.8 1.2 8.9
332 376 371.5 119.5 1.1 0.9 8.1
363 374 371.5 148.8 1.3 0.8 11.7
370 373 370 135 1.0 1.0 11.8
Downloaded from https://academic.oup.com/jge/article/3/3/207/5127598 by guest on 18 April 2025
Figure 3. Distribution map of amplification factors.
Figure 4. Seismic microzonation map of Rabat.
table 1 the assessed values of dominant periods for sites 53
and 112, located in this zone (figure 2) where the soil deposits
are thinner and the topography is relatively uniform. underlying the rock. This kind of site effect generally takes
The lowest values of dominant periods (less than 0.4 s) place in valleys recently filled with sediment.
were found in particular positions where the upper soil layers In order to illustrate the amplification phenomenon, we
are thinnest. The altitude of site 150 and 155, given in table 1, present in figure 3 the distribution of amplification factors over
allows us to interpret the lowest values of dominant periods the geotechnical map of Rabat. The values of amplification
assessed in these sites. The geographical position of these factors mainly range between 1 and 5, twenty values range
sites at low altitudes leads us to assume that the thickness of from 5 to 7 and only four values of amplification factors,
the sediment layer is very weak and therefore the value of estimated in some sites to be located in high topography,
dominant periods are very low with comparison to other sites exceeded 7.
of measurement. The transfer functions assessed at sites 363 and 370
(figure 3) show very high values of amplification factors.
These sites are situated on sandy soil and in high topography;
4.2. Distribution of amplification factors
their elevation taken from the topographic map of Rabat and
The amplification of the ground motion by the near surface their relative amplification factors are presented in table 1. The
layers due to the site effect is observed in two cases. The high values of amplification factors assessed at these sites can
first category of site effect is related to the shape of the local be explained by their location on a convex topography.
topography, for sites located on the crests of mountains or As we observe in figure 3, the two other high values of
hills, the seismic waves are focused due to the convexity amplification factors are obtained at sites 331 and 332. These
features. The second category of site effect is associated sites are located in high topography near ridge crests created
with the travelling of the seismic waves into soft material by the Bou Regreg River. Their geographical characteristics
210
Seismic site effect estimation in the city of Rabat
and their amplification factors, presented in table 1, lead and accentuated topography. The level of the amplification
us to attribute the high values of amplification to the shape becomes approximately uniform in the plain domain of
topography effect. investigated area. In general, both topography and thickness
of soil may have contributed in these site effects.
The seismic microzonation map established in this study
5. Seismic microzonation based on dominant periods contributes to the analysis of the seismic risk in the urban
zone of Rabat. The knowledge of the dynamic characteristics
The best way to present the results of this study is to establish a
of man made structures and the dominant periods at each
microzonation map of the studied area based on the estimated
site allow the assessment of the possibility of resonance
values of dominant periods in each site measurement. For
phenomenon between structures and the underlying soil
this purpose, we use a triangular interpolation to define six
column.
zones each one is characterized by a rank of dominant periods.
(figure 4).
References
Akil M 1980 Contribution à l’étude sédimentologique des
6. Conclusion formations littorales quaternaires de la région de rabat Thèse de
3 éme cycle Université Med. V. Rabat. Maroc.
Downloaded from https://academic.oup.com/jge/article/3/3/207/5127598 by guest on 18 April 2025
This experimental study allowed us to estimate the dominant Andrien M L 1974 Carte Géotechnique de la Région de Rabat.
periods and their relative amplification factors in more than 1/50.000, Editions du Service Géologique du Maroc Notes et
250 sites on the urban zone of Rabat. The analysis Mémoires no 238
and the comparison of dominant periods and amplification Bard P Y 1998 Microtremor measurements: a tool for site effect
estimation? ESG98: 2nd Int. Symp. on the Effects of Surface
factors with the geological and topographical information in Geology on Seismic Motion (Yokohama, Japan, December
the investigated area allowed us to infer many conclusions 1998)
concerning the seismic site response in the city of Rabat. Chavez-Garcia F J 1998 Site effets and microzonation in Acapulco
The geomorphologic and the geological setting of the Earth. Spectra 14 75–93
studied area shows that its surface topography has an Duval A M, Méneroud J P, Vidal S and Bard P Y 1996 Une nouvelle
Méthode d’évaluation de la réponse des sols aux séismes par
undulating shape and that the sedimentary deposits slope enregistrement du bruit de fond Bulletin des Laboratiores des
toward the Atlantic Ocean. The thickness of upper soil layer Ponts et Chaussées Réf. 4044 pp 75–90
consisting of loose deposits (sand and alluvium) decreases Field E H and Jacob K H 1993 The theoretical response of
from the borders of the Bou Regreg River in the north toward sedimentary layers to ambient seismic noise Geophys. Res.
the southern part of the investigated area. Lett. 20 2925–8
Lachet C and Bard P Y 1994 Numerical and theoretical
The observation of the distribution of dominant periods investigations on the possibilities and the limitations of
shows that the values of dominant periods are in general higher Nakamura’s technique J. Phys. Earth 42 377–97
on the hills bordering the Bou Regreg River. The values of Laouina A, Taı̈lassane M and Watfeh A 1994 Mise au point sur la
dominant periods clearly shift from the north to the south Géologie du plio-qauternaire et les formations superficielles de
away from the Bou Regreg River and parallel to the coastal la region de Rabat – Salé Revue de Géographie du Maroc 16 2
Nakamura Y 1989 A method for dynamic characteristics
line of Rabat. As a first approximation, we can consider in this estimations of subsurface using microtremors on the ground
study that the variation of dominant period values agree with surface Quaterly Report of Railway Technical Research
the surface geological conditions and confirm the relationship Institut, Tokyo vol 30 pp 25–33
between the dominant period and the thickness of the upper Nakamura Y 2000 Clear identification of fundamental idea of the
deposit layer. The small variability in site response over a Namamuras technique and its applications 12th World Conf. of
Earthquake Engineering (Auckland, New Zealand) paper 2656
short distance observed in some zones of the studied area can Perelman N and Zaslavsky Y 2001 Analysis of Seismic Signals in
be explained by the abundance of pockets filled by very soft the Spectral Domain (SEISPECT) (Israel: The Geophysical
sediments in the soil and by the undulating shape of the surface Institute of Israel)
topography of Rabat. Talhaoui A, Iben Brahim A, Aberkan M, Kasmi M and
On the other hand, the study of the effect of surface El mouraouah A 2004 Seismic microzonation and site effects
at al hoceima city, Morocco J. Earth Eng. 8 585–96
topography on the seismic response in the investigated area Wernli R 1977 Les formations plonctoniques de la limite
shows that the highest values of amplification factors are miopliocène dans les environs de Rabat Maroc. Ecologie. Geol.
obtained in sites situated in environments with high elevation Helv. Bâle. 70 143–91
211
