Soil Dynamics and Earthquake Engineering 43 (2012) 174–185

Contents lists available at SciVerse ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Probabilistic method to select calculation accelerograms based on uniform

seismic hazard acceleration response spectra
Antonio Morales-Esteban a,n, José Luis de Justo a, Francisco Martı́nez-Álvarez b, J.M. Azañón c
a
Department of Continuum Mechanics. University of Seville, Spain
b
Department of Computer Science. Pablo de Olavide University of Seville, Spain
c
Department of Geodynamics, University of Granada, Spain

a r t i c l e i n f o a b s t r a c t

Article history: A dynamic analysis of a structure requires the previous definition of the accelerograms and the
Received 8 March 2012 structure characteristics. The response of a structure subject to a seismic movement can be determined
Received in revised form by two methods: either using the accelerograms recorded near the site, or using visco-elastic response
19 June 2012
spectra. The first method should only be used for locations where many accelerograms have been
Accepted 8 July 2012
recorded, and needs a probabilistic calculation to ascertain the design accelerograms. The use of visco-
elastic response spectra is based upon the fact that the response spectrum is the soil movement
parameter better related to the structural response and is more adequate to obtain accelerograms in
regions where the number of records is insufficient. This is the most commonly used method as the
response of structures, in the elastic linear range, can be obtained as the superposition of a few modes
of vibration. A probabilistic method for selecting calculation accelerograms is presented in this paper.
First, the probabilistic hazard equation is solved. Based on the hazard curves obtained, the uniform
seismic hazard acceleration response spectrum (USHARS) is constructed for the location, according to
the type of soil and the required hazard level (exposure time and exceedance probability). Then,
calculation accelerograms are selected. Based on this methodology, real accelerograms, for a return
period of 975 years, have been obtained for San Pedro Cliff (Spain) at the Alhambra in Granada.
& 2012 Elsevier Ltd. All rights reserved.

1. Introduction
The dynamic response of a structure affected by a seismic
movement can be estimated by two methods. The first one consists
in defining a theoretical model of the structure and calculating its
dynamic response due to a given movement of its foundation. The
other method, which is approximated, consists in separating the
characteristic of the structure and the characteristics of the seismic
movement, represented by the response spectra.
To keep a balance between economy and security, current
seismic regulations generally accept that, for large earthquakes,
structures might suffer non-linear deformations without collap-
sing. The non-linear dynamic analysis more used is the calcula-
tion, step by step using accelerograms. This procedure can be
used at the linear and non-linear range. A structural analysis of all
the accelerograms considered must be worked out, and a calcula-
tion envelope must be obtained. The dynamic calculation envel-
ope must be later combined with the static loading envelope.
This procedure implies a significant work; it is only used for the
n
Corresponding author. Tel.: þ34616360273; fax: þ 34954541007.
E-mail address: ame@us.es (A. Morales-Esteban).
design of very important structures and only in the last stages of
the calculation procedure. In some regions, with a vast history
of large earthquakes, such as Japan and California, a wide network
of recording stations is available and provides many records for
large earthquakes, for different types of soils and for a wide range
of distances. In regions of minor seismicity, the network of
recording stations is not so wide, or is not old enough, so that
the number of records is insufficient. For the analysis of minor
seismicity activity regions, records from other regions are used, or
artificial accelerograms are generated.
Between 5 and 10, real or artificial, accelerograms, must be
obtained, and must be scaled to a level of severity. The most
commonly used method consists of scaling the seismic peak
acceleration up to a predetermined probabilistic level. However,
the potential damage an earthquake can produce is not only a
function of the peak acceleration; there are many alternatives
such as the Arias intensity and the spectral intensity of Housner
[24,34] more related to the potential damage.
The use of visco-elastic response spectra, based on the answer
of structures at the elastic linear range can be estimated by the
superposition of a few modes of vibration and is the most
commonly used method due to its simplicity and appropriate
accuracy.

0267-7261/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.soildyn.2012.07.003
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185
A method to estimate the uniform seismic hazard acceleration
response spectrum (USHARS) is presented, based on it calculation
accelerograms can be selected.
2. Fundamentals
This section exposes the fundamentals that support the meth-
odology applied to calculate the seismic hazard. First, seismic
hazard is defined. Then, the methods used to calculate the seismic
hazard are presented and discussed.
Seismic hazard of a location can be defined as the probability
to equal or exceed a parameter of the soil movement, produced by
the earthquakes that occur on the influence area, during a
specified lapse of time. To unify criteria, UNESCO proposed the
commonly accepted definition, given by UNDRO [43]. Hazard (H)
is defined by a probability function of the characteristic para-
meter of the soil movement (S) at the location (x) according to:
H ¼ P½SðxÞ ZS0 ; t
where P represents the probability of exceeding a threshold value
(S0) of the characteristic parameter of the soil movement, during
the time (t), in years.
Seismic hazard is defined as a probabilistic function, where the
value of the soil movement, that can be considered dangerous,
must be specified and the time of exposure must be established.
The following methods to estimate the seismic hazard may be
used: deterministic methods, probabilistic methods, historical
and deductive methods and other methods.
2.1. Deterministic methods
The first methods to establish the seismic hazard of an area or
a location were deterministic. The characteristic parameter of the
soil movement used is usually the peak ground acceleration
(PGA). The deterministic methods assume the hypothesis that
the seismicity is stationary, considering that earthquakes in the
future will be similar to those in the past. The upper limit of the
movement, expressed as the maximum value of the parameter, is
estimated. These earthquakes can be real earthquakes that in the
past affected the location, or can be deduced from the seismic and
tectonic characteristics of the area. The deterministic method can
be divided into zoned or not-zoned, in function of how the
seismicity distribution is considered.
The calculation procedure is [39]:
1. Definition of the influence area of the location, and identifica-
tion of the seismic sources or faults within it.
2. Estimation of the largest earthquakes that have occurred in the
influence area or at any of the source areas.
3. Estimation of the seismic parameter at the location, caused by
the maximum potential earthquakes of every area or of the
whole area.
4. Determination of the hazard at the location, taking the max-
imum value generated by the influence areas. Hazard is
defined by the upper limit of the movement at the location.
This method presents some advantages and disadvantages. Its
main advantage is its easy application. It defines earthquakes that
happened in the past and supposes that similar earthquakes will
happen in the future. Normally, these methods define the most
unfavourable seismic situation (the largest earthquake in the
closest source). However, the probability that these earthquakes
will happen in the future is, generally, unknown. The determi-
nistic method estimates the largest earthquake that can affect
the location, while the rest of earthquakes are not considered.
175
The sources are characterised by the largest earthquake, and not
by its recurrence law.
2.2. Probabilistic methods
Later, the probabilistic methods were defined by Cornel [14].
The probabilistic methods sum up the contribution of all the
possible earthquakes that can affect a location, and consider
recurrence laws for them. As a result, the probability of exceeding
every value of a parameter of the soil movement expected at the
location, during a period of time, is estimated. The hazard is
represented by probability curves. The development of the soil
strong motion prediction models allowed the construction of
design spectra, where the probability to exceed all the spectra
ordinates is considered (and not only peak acceleration). The
uniform hazard spectra represent the predicted values for a
return period of all the spectral ordinates [42]. These methods
are classified into parametric and non-parametric, according to
ð1Þ
the statistical distribution adopted.
2.2.1. Non-parametric methods
These methods analyse the hazard according to extreme value
distribution functions. The most used were defined by Gumbel [23].
The method is based on the following steps:
1. Definition of the area of influence around the location.
2. Calculation of the values of the seismic parameter at the
location, applying attenuation laws to the values of the
parameter, that reflect the seismicity of the area during the
period of time considered.
3. Adjustment to a distribution of extreme values of the random
parameter, defined with the values of the estimated para-
meter, and estimation of the distribution coefficients.
4. Estimation of the probability of exceeding the extreme value,
during the time considered, calculating, thus, the probability.
2.2.2. Parametric methods
The methodology was initially proposed by Cornell [14]. The
method is based on the existence of different seismogenic areas.
First, the influence area is divided into seismogenic areas and the
seismicity of every area is adjusted according to a recurrence law.
Later, the contribution of all the sources is added to obtain a
probability function that represents the hazard at the location.
The uniform hazard spectrum is a very useful tool for Seismic
Engineering, mainly for structures with various degrees of free-
dom, for which a uniform hazard level is demanded [5]. Recently,
methods to evaluate seismic hazard that combine probabilistic
and deterministic characteristics have been proposed. Most of
these methods are specially developed for the design of seismic
structures at specific locations [40,35,36,32].
Currently, specially for academic studies, deterministic meth-
ods [13,30] and probabilistic methods are used [41,21]. Although
some regulations suggest the use of both methodologies [25],
generally the use of probabilistic methods is prioritised [9],
particularly for nuclear security [20,33].
2.3. Historical and deductive methods
Historical methods do not make hypotheses about the char-
acteristics of the faults and the seismic parameters that might
cause earthquakes [45]. However, deductive methods make
hypotheses about the origin of earthquakes.
176 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185

2.4. Other methods
Traditional methods for the seismic hazard calculate the
probability of exceeding an established level of the soil movement
at the location. On the other hand, these methods do not assess
the joint probability to exceed those levels of the soil movement
at multiple locations.
Aforementioned, probabilistic methods allow calculating uni-
form hazard spectra, which are better related to the structural
response. For this reason, a probabilistic method has been used in
this paper.
3. Methodology
In this section the methodology applied to select calculation
accelerograms is presented. First, the probabilistic method is
described. Then, the probabilistic method is itemised into its four
parts: the seismicity model, the recurrence model, the attenua-
tion law and the probabilistic hazard equation. Later, the prob-
abilistic hazard equation is solved and the method to obtain the
USHARS is described. Finally, the equations to select calculation
accelerograms from the USHARS are described.
3.1. Probabilistic method to estimate seismic hazard
Seismic hazard is represented by a hazard function (H) that
indicates the characteristic parameter of the soil movement (S),
according to the following equation:
HðS0 ; tÞ ¼ PðS ZS0 ; tÞ
where P(SZS0;t) is the probability that the characteristic para-
meter of the soil movement will exceed a threshold value, S0, at
least once during the time, t.
The arrival of earthquakes to a location is assumed to follow a
Poisson’s stationary process [14,45]. Under this hypothesis, the
hazard function can be expressed as follows:
HðS0 ; tÞ ¼ 1elðS0 Þt
In this case, the seismic sources and the attenuation laws of the
soil movement, from the source to the location, must be analysed.
The steps to follow with this method were proposed by Cornell [14]:
1. Seismicity model: definition of the seismogenic areas that can
affect the site. If the seismicity can be considered homogenous
in the whole seismogenic area, a unique seismicity source of
global influence can be defined.
2. Recurrence model: the recurrence model in every seismogenic
area must be defined. If it is admitted that the seismicity is
randomly distributed and it adjusts to the Gutenberg–Richter
law with upper truncation, the parameters of the law (a and b)
are characteristics of the model. Moreover, for every area
maximum and minimum magnitudes are defined that estab-
lish the validity limits of the model.
3. Attenuation law: attenuation laws to obtain the selected
parameter in function of the distance must be determined in
order to evaluate the seismic hazard. The application of these
laws over the seismicity of every area, represented by its
recurrence law, allows obtaining the result over the site.
4. Probabilistic hazard equation: the estimation of the total
hazard is obtained adding up the probabilities obtained by
the result of all the areas that affect the site.
n 
X 
H¼ 1etl ð6Þ
i¼1
where l is the annual rate of earthquakes, occurring in any area
that produces a parameter of the soil movement superior to the
reference one at the studied site, n is the number of areas and t is
ð2Þ the period of time considered, in years.
The uniform seismic hazard response spectra are those that
have the same probability of being exceeded in all the periods,
obtained with the method proposed by Cornell [14].
3.1.1. Seismicity model
Seismicity can be assimilated to a process of punctual events
ð3Þ that result from the relaxation of stress that acts over an area.

where l(S0) is the annual rate of times that the parameter (S0) has Table 1
been exceeded at the location. Seismogenic areas of Spain and Portugal.
To conduct a study of the seismic hazard, a database of
earthquakes that can affect the location must be provided. If the Area Description
value of the characteristic parameter of the soil movement, in
1 Granada basin
every one of these earthquakes, is known and the database is 2 Penibetic area
complete, even with the largest earthquakes that can affect the 3 Area to the East of the Betic system
location, the annual rate of exceedance could be calculated 4 Quaternary Guadix-Baza basin
according to the following formula: 5 Area of moderate seismicity to the North of the Betic System
6 Area of moderate seismicity including the Valencia basin
1X 7 Sub-betic area
lðS0 Þ ¼ dðSk S0 Þ ð4Þ 8 Tertiary basin in the Guadalquivir depression
tc k
9 Algarve area
where tc is the duration of the database, Sk is the value of the 10 South-Portuguese unit
11 Ossa Morena tectonic unit
characteristic parameter of the soil movement during the k-th
12 Lower Tagus Basin
earthquake of the database, and d is Heavisides function: 13 West Portuguese fringe
( 14 North Portugal
1-Sk S0 Z 0
dðSk S0 Þ ¼ ð5Þ 15 West Galicia
0-Sk S0 o 0 16 East Galicia
17 Iberian mountain mass
Eq. (4) provides a good estimation of l(S0) only if all the 18 West of the Pyrenees
earthquakes, that can affect the location, have been presented 19 Mountain range of the coast of Catalonia
20 Eastern Pyrenees
various times during the period of time the database covers. This
21 Southern Pyrenees
implies that the database should have a very long duration, 22 North Pyrenees
probably of thousands of years [19]. 23 North-Eastern Pyrenees
If an instrumental parameter is taken as characteristic of the 24 Eastern part of Azores–Gibraltar fault
soil movement, the annual rate of exceedance l(S0) cannot be 25 North Morocco and Gibraltar field
26 Alboran Sea
obtained by means of (4) as there are no databases of enough 27 Western Azores–Gibraltar fault
duration.
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185 177

For its study, the spatial distribution of the earthquakes and its
occurrence, according to time, must be known.
The areas or seismic sources, named seismogenic areas, are
lithosferic volumes associated to certain tectonic characteristics,
where earthquakes with similar tectonical origin are supposed to
happen. The process of earthquake generation is spatially and
temporarily homogenous in every area. A seismogenic source,
generally, is related to one or a few active faults [4] and its
location and geometry is of great importance to evaluate its
hazard. The task of delimiting seismogenic areas depends on the
available information of the studied region. When sources corre-
spond to inter-plate areas, are associated to large active faults and
probably superficial, it is much easier to identify them. Contrarily,
when the sources correspond to intra-plates, that are not asso-
ciated to active faults, the task can be very difficult.
When the seismogenetic sources are vast and it is difficult to
delimit all the included active faults, it is necessary to establish
seismotectonic delimitation, that is, a subdivision of the territory
under study in areas with a homogenous seismic behaviour from
a tectonic point of view.
The model used in this text is based on the seismogenic areas
defined by Morales-Esteban et al. [38]. The twenty seven areas
established for the Iberian Peninsula are based on tectonic,
geological, seismic and gravimetric data. The twenty seven areas
are described in Table 1 and Fig. 1 (Seismogenic areas that might
affect Spain and Portugal. Dots represent earthquakes of moment
magnitude between 3.0 and 4.0, circles between 4.0 and 5.0, and
solid circles of magnitude larger than 5.0).
3.1.1.1. Seismicity in the Iberian Peninsula. Currently, the convergence
directed NWSE, between Africa and Eurasia, throughout this part of
the limit between plates, is responsible for the deformation of the
crust of the Iberian Peninsula, the Maghreb, and the adjacent coastal
areas of the Mediterranean and the Atlantic [3,31,26]. The plates
limit is not homogenous, with consecutive oceanic and continental
areas in contact and progressive changes in the stress direction. The
area corresponding to the Iberian Peninsula and the northwest of
Africa can be considered the most complicated contact area, with a
moderate seismicity in relation to the magnitude of the earthquakes.
This area is surrounded at both sides by a frequent seismic activity
with very large earthquakes [22,44]. The seismic activity of this area
is of great importance as large earthquakes have been produced in
the past such as the Lisbon 1755 earthquake. The coexistence of
compressive and extensive tectonics, and the interference of the
Iberian micro-plate [2,12], makes the understanding of the area very
complicated and many issues of the tectonic structure of the area are
prompting discussion among experts. The regional seismicity is

12°W 6°E
9°W 3°E 45°N
45°N 6°W 3°W 0°W

23
16
18 22
15 21 20
42°N 42°N

17 19
14

13

39°N 12 39°N

5 6
11

10 4
8 7 3
9
1
2
27
36°N 36°N

24 26

25

0 100 200 300 km

33°N 33°N
12°W 6°E
9°W 0° 3°E
6°W 3°W

Fig. 1. Seismogenic areas of Spain and Portugal. Dots represent earthquakes of moment magnitude between 3.0 and 4.0, circles between 4.0 and 5.0, and solid circles of
magnitude larger than 5.0.
178 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185

diffuse and is not clearly aligned with the current limit between the less than 5. Large earthquakes are separated by long periods of
plate of Eurasia and Africa, at the south of the Iberian Peninsula [11]. time [10]. Many of the earthquakes are located at the east of the
The seismic activity extends to far away inter-plate areas, such as the Gibraltar Arch and spread over a diffuse area, of approximately
northeast and the centre of the Iberian Peninsula. 500 km wide, centred at the Alboran Sea, containing parts of the
The seismicity in Spain and Portugal is characterised by the southeast of Spain, the north of Morocco and Algeria. The limit
occurrence of moderate earthquakes, with a magnitude generally between the plates of Africa and Eurasia is, currently, a matter
under investigation [26]. At the west of Gibraltar, most earth-
quakes happen in the southern coast of Portugal next to the limit
between the plates of Azores and Gibraltar. Other seismic sources
Table 2
Axis orientation, regime of stress, maximum magnitude and predominant type of include the northwest of Spain and the Pyrenees. Earthquakes
faulting for the seismogenic areas of the Iberian Peninsula. rarely happen in other location. This seismicity so widespread
happens over geo-tectonic areas with different structure and
Area Regime Axis orientation Predominant type of faulting Mmax
rheology. At the Iberian Peninsula, at least three different tectonic
1 Compressive P EN1701E Normal 7.0
regimes can be differentiated: stable blocks, alpine mountain
2 Compressive P EN1701E Normal 6.5 belts and extension basin.
3 Compressive P EN1701E Strike-slip 7.0 The work by Morales-Esteban [37] has studied the faulting
4 Compressive P EN1701E Strike-slip 6.0 mechanism for the seismogenic areas of the Iberian Peninsula
5 Compressive P EN1701E Normal 5.5
proposed by Morales-Esteban et al. [38]. In Table 2 and Fig. 2
6 Compressive P EN1701E Strike-slip 6.5
7 Compressive P EN1701E Normal 6.0 (Axis orientation, stress regime, maximum magnitude and type of
8 Compressive P EN1701E Thrust 6.5 faulting for the seismogenic areas defined by Morales-Esteban
9 Compressive P EN1701E Thrust 6.0 et al. [38] for the Iberian Peninsula) the orientation, for the 27
10 Compressive P EN1701E Thrust 5.5
seismogenic areas, of the main fault axes is shown.
11 Compressive P EN1701E Thrust 5.5
12 Compressive P EN1351E Thrust 7.0
13 Compressive P EN1351E Thrust 5.5 3.1.1.2. Seismicity at San Pedro cliff in Granada. Granada basin
14 Compressive P EN1351E Thrust 5.5 presents several sets of faults, most notably those with E–W and
15 Compressive P EN1351E Strike-slip 5.5
NW–SE orientations. Conspicuous NW–SE faults are present in the
16 Compressive P EN1351E Strike-slip 6.0
17 Compressive TEN451E Strike-slip 5.5 eastern part and the eastern limit of the basin. These faults are
18 Compressive P EN1701E Thrust 6.0 normal, mostly with a NW–SE orientation, and dipping towards the
19 Compressive P EN01E Normal 5.5 SW. These NW–SE faults cross-cut and displace previous E–W
20 Compressive P EN01E Thrust 6.0 faults, defining the main subsiding areas of the central and
21 Compressive P EN1701E Thrust 5.5
22 Compressive P EN1701E Thrust 5.5
eastern part of the basin. Fig. 3 (Annual rate of earthquakes by
23 Compressive P EN1701E Thrust 6.0 square kilometre and b-value for the seismogenic areas defined by
24 Compressive P EN1701E Thrust 7.0 Morales-Esteban et al. [38] for the Iberian Peninsula) shows the
25 Compressive P EN1701E Normal 5.5 b-value and the annual rate of earthquakes by square kilometre for
26 Compressive P EN1701E Strike-slip 6.5
the 27 seismogenic areas. It can be observed that the annual rate of
27 Compressive P EN1601E Thrust 8.5
earthquakes in the Granada basin (area 1) is clearly the highest.

Fig. 2. Axis orientation, stress regime and type of faulting for the seismogenic areas defined in [37] for the Iberian Peninsula.
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185 179

However, the b-value is an average (a low b-value implies that the
probability of large earthquakes to happen is high). This means that
earthquakes in Granada are very frequent (clearly the most active
seismogenic area) although, normally, not of large magnitude.
3.1.2. Recurrence model
The seismicity in every seismogenic area is randomly distrib-
uted and it adjusts to the Gutenberg–Richter law. In seismic
hazard studies, the Gutenberg–Richter law must be truncated,
with upper and lower limits, to consider the magnitude, Mmax, of
the largest earthquake that can occur at the source, and to avoid
considering earthquakes of magnitude less than Mmin, respec-
tively. The probability density function of magnitude for
Gutenberg–Richter law is:
to the source, and sometimes, with some other variables. After the
magnitude, the distance to the source and the type of soil at the
location, the most used parameter, for attenuation laws, is the
type of fault. From a database of soil movements an attenuation
law can be obtained through a regression analysis.
The characteristic parameter of the soil movement, used in this
text, is the spectral acceleration (SA). SA is the most suitable
parameter to select accelerograms because it is the soil move-
ment parameter better related to the structural response.
The equations for the estimate of ground motions from
shallow crustal earthquakes from Morales-Esteban [37] have been
used. These coefficients were obtained from strong motions
recorded in Europe and the Middle East which are consistent
with the location under study (the Iberian Peninsula). Douglas

ebðMMmin Þ
f ðMÞ ¼ b ð7Þ Table 3
1ebðMmax Mmin Þ
Annual rate of earthquakes, maximum magnitude and b-value for the seismogenic
The seismicity in every seismogenic area is defined by the areas that affect Spain and Portugal.
following parameters:
Area b-value Annual rate of Surface Anual rate/ surface Mmax
earthquakes (km2) (km2)
1. The maximum and minimum magnitude.
2. The annual rate of earthquakes occurrence between 1 1.41 5.14 3835 1.34E-03 7.0
Mmax and Mmin. 2 1.18 7.82 13,979 5.59E-04 6.5
3. The b-value of the Gutenberg–Richter law (b ¼b/log e). 3 1.29 4.36 13,251 3.29E-04 7.0
4 1.27 2.26 11,957 1.89E-04 6.0
5 1.62 0.87 7066 1.24E-04 5.5
The maximum magnitude in every seismogenic area has been 6 2.17 1.38 9735 1.42E-04 6.5
determined by Morales-Esteban et al. [38] from seismic and 7 1.51 4.32 13,954 3.10E-04 6.0
tectonic considerations. The minimum magnitude in all areas is 8 0.92 1.47 22,228 6.63E-05 6.5
9 1.20 0.77 6371 1.21E-04 6.0
4.5. Lower earthquakes are not considered dangerous. Normally,
10 2.33 2.56 15,717 1.63E-04 5.5
seismologists only consider earthquakes of magnitude equal or 11 1.44 2.35 27,694 8.47E-05 5.5
larger to 5.0. However, not very large earthquakes but close to the 12 1.01 0.50 9803 5.08E-05 7.0
source are known to have caused damage. Moreover, earthquakes 13 1.29 1.41 13,029 1.09E-04 5.5
in the Iberian Peninsula are known to be of moderate magnitude 14 1.40 1.75 26,049 6.71E-05 5.5
15 2.07 4.02 22,597 1.78E-04 5.5
and magnitudes over 5.0 are uncommon. The b-value, the
16 1.54 2.87 15,475 1.85E-04 6.0
maximum magnitude and the annual rate of earthquakes can be 17 1.75 1.07 26,993 3.96E-05 5.5
obtained from Table 3. 18 1.54 0.58 15,738 3.70E-05 6.0
19 1.90 1.18 16,032 7.35E-05 5.5
20 1.63 1.68 10,622 1.58E-04 6.0
3.1.3. Attenuation law 21 1.53 2.63 19,946 1.32E-04 5.5
Equations for the estimation of the parameters of the soil 22 1.48 12.32 22,383 5.51E-04 5.5
movement are a basic tool for the calculation of the seismic 23 1.46 2.06 4301 4.78E-04 6.0
hazard. 24 0.96 13.55 46,329 2.92E-04 7.0
25 0.96 5.85 24,600 2.38E-04 5.5
Attenuation laws are functions that relate a parameter of the
26 1.14 18.21 48,669 3.74E-04 6.5
soil movement (peak acceleration, spectral acceleration, relative 27 0.70 15.16 38,955 3.89E-04 8.5
velocity response spectra, etc.) with the magnitude, the distance

2.50
10
615
2.00
19
17
5 20
18 21 16 7
b 1.50 11 23 22 1
14
13 4 3
9 2
26
1.00 12 25 24
8
27
0.50

0.00
0.00E+00 2.00E-04 4.00E-04 6.00E-04 8.00E-04 1.00E-03 1.20E-03 1.40E-03 1.60E-03
Annualrate/Area(km2)

Fig. 3. Annual rate of earthquakes by square kilometre and b-value for the seismogenic areas defined in [37] for the Iberian Peninsula.
180 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185

[17] has showed that recorded strong ground motion in the
Caucasus region, central Italy, Friuli, Greece and south Iceland
shows little evidence for regional differences. However, he found
evidence for regional differences in ground motions between
Europe and California [18].
The magnitude scale used is the moment magnitude, Mw, [29].
The use of the moment magnitude avoids the saturation of the
traditional magnitude scales for the larger earthquakes and,
consequently, is considered a better measure of the real size of
an earthquake [6]. Earthquakes with Mw o4.5 have been excluded
in order to avoid earthquakes that are unlikely to be of engineer-
ing significance. Ambraseys et al. [1] in its regression analysis for
Europe, only consider earthquakes of magnitude larger than 5.0.
However, the minimum in this study has been reduced to 4.5, as
the seismicity in the Iberian Peninsula is moderate and few
earthquakes exceed a moment magnitude of 5.0.
Fault distance has been defined as the minor distance to the
surface projection of the fault as proposed by Joyner and Boore [27]
also known as Joyner–Boore distance. Douglas [16] has demon-
strated that using the distance to the rupture does not reduce the
standard deviation for ground motion prediction equations.
The scheme proposed by Bommer et al. [7] has been used
to classify the local geology. This scheme uses the average
shear wave velocity, estimated at a depth of 30 m (Vt30). Therefore
four site classes have been used: rock, Vt30 4750 ms  1; hard soil,
360 ms  1 oVt30 r750 ms  1; soft soil, 180 ms  1 oVt30 r
1 1
360 ms and very soft soil, Vt30 r180 ms .
Currently there are many criteria to classify the faulting
mechanisms Boore et al. [8] which can lead to different classifica-
tions for the same earthquake [7].
The coeficients to obtain the SA, for the 25 periods calculated
and for the 0, 2, 5, 10 or 20 per cent damping, can be obtained
from Morales-Esteban [37]. The predominant type of faulting for
every seismogenic area has been considered for the attenuation
laws according to Table 2.
3.1.4. Probabilistic hazard equation
It is admitted that the arrival at the site of earthquakes that
exceed the reference value, log S0, follows a Poisson stationary
process, defined by Gutenberg–Richter law of constant li:
Z Mmax
li ¼ ui Pðlog S Z log S0 =M,DÞf ðMÞdM ð8Þ
Mmin
The seismic rate of the punctual source is ui and f(M) is the
probability density function of magnitude (Eq. (7)).
If N punctual seismic sources hit simultaneously the site, the
rate l of arrivals at the location that exceed the reference value
log S0 is:
X
N
l¼ li ð9Þ
i¼1
The probability of exceeding the reference value log S0 during
a time t caused by the simultaneous action of N punctual seismic
sources is:
PðlogS Z logS0 ; t Þ ¼ 1elt ð10Þ
Its return period can be obtained from:
1 t
T¼ ¼ ð11Þ
l lnð1PðlogS Z logS0 ; tÞÞ
Eq. (8) cannot be applied to the hazard calculation as the
seismogenic areas have been modelled as areas and not as
punctual seismic sources. To solve this problem, the seismogenic
areas are divided into elements small enough to be assimilated to
punctual seismic sources (Fig. 4. Scheme of division of a seismo-
genic area into N punctual areas, through an orthogonal mesh).
3.2. Uniform seismic hazard acceleration response spectrum
A computer programme that divides the seismogenic areas
into punctual seismical sources that affect the location that
integrates numerically Eq. (8) and calculates the exceedance
probability with Eq. (10) has been developed.
With this method, the probability of exceeding a value of the
acceleration response spectra for an established time of exposure
can be calculated. Following this methodology a plot of the
seismic hazard is obtained (Fig. 5).
If this process is repeated for various periods of the structure
and, for every seismic hazard plot, the value of the spectrum for
the same exceedance probability is obtained, and a uniform
seismic hazard acceleration response spectrum (USHARS) can be
obtained, point by point, as shown in Fig. 6 (Scheme of construc-
tion of a uniform seismic hazard response spectrum from the
seismic hazard plots).

1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
8 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
7 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4
1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4
2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4
3
1 1 1 1 1 1 1 1 1 1 1 1 4 4
1 1 1 1 1 1 1 1 1 1 4 4
26 1 1 1 1 1 1 1 1 1 4
2 2 2 2

Fig. 4. Scheme of division of a seismogenic area into Nc punctual areas, through an ortogonal mesh.
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185 181

3.3. Selection of calculation accelerograms
The procedure to select design accelerograms is:
1. The time of exposure of the structure is established, according
to the hazard level.
2. The admitted exceedance probability is established, normally a
5–10%, according to the hazard level.
3. The USHARS for the location is calculated according to the type
of soil and the required hazard level (time of exposure and
probability of exceeding).
4. Records from the accelerograms database, registered in the
same type of soil of the location, are examined. The factor of
scale, f, between the logarithm of the USHARS calculated and
the logarithm of the response spectrum which corresponds to
the real spectrum, that minimises the standard deviation, s, is
calculated.

So, if SR is the response spectrum correspondent to the real

register, and SC is the calculated response spectrum, the standard
deviation is:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2
logðf USR ÞlogðSC Þ
s¼ ð12Þ
25
The scale factor that minimises the standard deviation is:
SlogSC SlogSR
f¼ ð13Þ
25
The sum is extended to the 25 periods for which the USHARS
has been calculated. The records for which the standard deviation
is minor are selected.

4. Application to San Pedro cliff

The Alhambra in Granada is one of the most important

Fig. 5. Comparison for different exposure time of the seismic hazard plot of the
acceleration response spectrum. Granada for a 0.50 s period, in rock, a 5% relative
dumping.
national monuments in Spain. This monument, a World Heritage
site, is located on the top of a red hill that dominates a plain, the
Granada basin, where most of the city is placed. One of the most
important rivers of the region, River Darro, flows into the basin
and is situated on the western part of the city. The Alhambra’s
walls are close to the escarpments generated by the incision of
this river. Slope instability of the escarpments on this side of the
Alhambra hill has been a critical problem since the construction
of this palace. In this area, San Pedro Cliff (Fig. 7), a dihedral
65.5 m high, is the steepest escarpment of the Alhambra hill. This
eroding cliff reaches to 23.8 m from the Alhambra palace wall.

T1 T2 T3

P0 P0 P0

S1 S2 S3

S2
S3
S1

T1 T2 T3

Fig. 6. Scheme of construction of a uniform seismic hazard response spectrum from the seismic hazard plots.
182 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185
Fig. 7. South view of San Pedro Cliff, showing to the right the fault line scarp. Above stand the Alhambra walls and, at the foot, Darro River and Albaicin houses.
Fig. 8. Comparison of the seismic hazard plot of the acceleration response spectra
of period 0.50 s for Granada for different type of soils, for a 5% relative damping
and a time of exposure of 50 years.
Retreat of this cliff has occurred through superficial slab falls
mainly induced by the floods of the Darro River, the loosening
produced by the extensional tectonic regime, erosion, seepage
coming from the Alhambra palace and earthquakes.
As an example of the methodology presented in this paper,
calculation accelerograms for San Pedro Cliff are selected. First,
seismic hazard curves for the SA for San Pedro Cliff have been
obtained. Fig. 8 (Comparison of the seismic hazard plot of the
acceleration response spectra for different soil types. Site Gran-
ada; period 0.50 s; relative damping 5%; time of exposure
50 years) shows the curves for a period of 0.5 s, for a 5% relative
damping and a time of exposure of 50 years, as a function of the
soil type. Fig. 9 (Comparison for different damping ratios of the
seismic hazard plot for the acceleration response spectra. Site
Granada and rock foundation; period 0.20 s; time of exposure 50
years) compares the seismic hazard curves, as a function of the
relative damping, for the acceleration response spectra of period
Fig. 9. Comparison for different damping ratios of the seismic hazard plot for the
acceleration response spectra. Period 0.20 s, in rock for a time of exposure of
50 years for Granada.
0.20 s and for a time of exposure of 50 years. Fig. 5 compares the
seismic hazard curves as a function of the time of exposure
(50 and 100 years) in rock, for a period of 0.50 s and a relative
damping of 5%.
Secondly, the type of soil at the location must be known. A site
investigation was conducted by Justo et al. [28]. The following
layers appear from top to bottom in the geological profile:
(1) Dense conglomerate. Vs ¼800 m/s (transverse wave velocity).
(2) Very dense conglomerate. Vs ¼960 m/s.
(3) Moderately dense conglomerate. Vs ¼800 m/s.
(4) Very dense, gravelly and sandy conglomerate. Vs ¼1150 m/s.
(4.a) One metre thick clay layers, interspersed in layer (4).
Vs ¼800 m/s.
Talus appears at the foot of the slope, composed of quartzose
and phyllitic blocks, gravel and sand, with predominance of the
sand fraction.
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185 183

The whole Cliff lies over the very dense, gravelly and sandy
conglomerate layer. Therefore, the transverse wave velocity at
30 m (Vs30) is 1150 m/s. The transverse wave velocity was
obtained through down-hole and cross-hole tests [28].
Thirdly, the methodology described to obtain the USHARS has
been applied to San Pedro Cliff, for rock (Vs Z750 m/s), for a 5%
probability of being exceeded and a time exposure of 50 years,
which is equivalent to a return period of 975 years.
Finally, accelerograms from the [15] European earthquake
database, that are available on-line, recorded in rock, have been
compared with the USHARS, obtained for San Pedro Cliff. The
accelerograms selected are those whose standard deviation is
minor. A total of 10 accelerograms, whose standard deviation is
lower to 0.13, have been obtained. The European Strong Motion
database provides the fault distance, which is the criteria used to
characterise the distance from selected site to source in this
paper. The fault distance for the 10 selected accelerograms varies
from 11 km (record 365) to 65 km (record 5826).
Figs. 10 and 11 represent the USHARS for San Pedro Cliff
and the response spectra of the real earthquakes with better
adjustment. The scaled spectrum that minimises the standard
deviation has also been plotted. The information about the
records that minimise the standard deviation is shown in
Tables 4 and 5.
5. Conclusions
A method to select real accelerograms for dynamic calculation
analysis is presented in this paper. This method has been
designed as a general method for any location of the Iberian
Peninsula.
The visco-elastic response spectrum has been used to select
design accelerograms which is consistent with the number of
accelerograms available for the Iberian Peninsula. It is known as
well that a response spectrum is the soil movement parameter
better related to the structural response.
A parametric probabilistic method, based on seismogenic
areas, has been used to calculate the seismic hazard. The meth-
odology presented in this paper allows the use of other methods

Fig. 10. Seismic hazard acceleration response spectra for San Pedro Cliff. Exceedance probability 5%; relative damping 5%; rock foundation; time of exposure 50 years.
Comparison between the calculated uniform acceleration response spectrum, the spectrum corresponding to record 128, whose fault distance is 19 km, from the catalogue
and this spectrum scaled to minimise the standard deviation.

Fig. 11. Seismic hazard acceleration response spectra for San Pedro Cliff. Exceedance probability 5%; relative damping 5%; rock foundation; time of exposure 50 years.
Comparison between the calculated uniform acceleration response spectrum, the spectrum corresponding to record 990 128, whose fault distance is 13 km, from the
catalogue and this spectrum scaled to minimise the standard deviation.
184 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185

Table 4
Information about the records whose typical deviation is minor in relation to the
uniform seismic hazard acceleration response spectra for San Pedro Cliff for an
exceedance probability of 5%, a time of exposure of 50 years, rock foundation and a
relative damping 5%.
Table 5
Information about the records whose typical deviation is minor in relation to the
uniform seismic hazard acceleration response spectra for San Pedro Cliff, an
exceedance probability of the 5%, a time of exposure of 50 years, rock foundation
and a relative damping of 5%.

Earthaquake record 128 Earthaquake record 5826

Seismical data Seismical data
Earthquake: Friuli (Northern Italy) Earthquake: Strofades (Jonic Sea)
Date: 9/15/1976 Date: 11/18/1997
Magnitude: 6.0 Mw Magnitude: 6.6 Mw
Record data Record data
Station: Robic (Slovenia) Station: Kyparrisia-Agriculture Bank (Greece)
Type of soil: Rock Type of soil: Rock
Fault distance: 19 Fault distance: 65
f 1.2 f 1.13
s 0.016 s 0.129
SAmax(m/s2) 3.96 SAmax(m/s2) 2.76

Earthaquake record 201 Earthaquake record 6265

Seismical data Seismical data
Earthquake: Montenegro (Adriatic Sea) Earthquake: Southern Iceland
Date: 4/15/1979 Date: 6/17/2000
Magnitude: 6.9 Mw Magnitude: 6.5 Mw
Record data Record data
Station: Dubrovnik-Pomorska School (Croatia) Station: Burfell Hydroelectric Station (Iceland)
Type of soil: Rock Type of soil: Rock
Fault distance: 61 Fault distance: 25
f 1.15 f 1.16
s 0.087 s 0.076
SAmax(m/s2) 2.68 SAmax(m/s2) 2.58

Earthaquake record 361 Earthaquake record 6270

Seismical data Seismical data
Earthquake: Umbria (Center of Italy) Earthquake: Suthern Iceland
Date: 4/19/1984 Date: 6/17/2000
Magnitude: 5.6 Mw Magnitude: 6.5 Mw
Record data Record data
Station: Nocera Umbra (Italy) Station: Ljosafoss Hydroelectric Station (Iceland)
Type of soil: Rock Type of soil: Rock
Fault distance: 19 Fault distance: 32
f 1.23 f 1.23
s 0.028 s 0.026
SAmax(m/s2) 7.49 SAmax(m/s2) 2.06

Earthaquake record 365 Earthaquake record 6331

Seismical data Seismical data
Earthquake: Lazio Abruzzo (Southern Italy) Earthquake: Southern Iceland, aftershock
Date: 5/7/1984 Date: 6/21/2000
Magnitude: 5.9 Mw Magnitude: 6.4 Mw
Record data Record data
Station: Atina (Italy) Station: Flagbjarholt (Iceland)
Type of soil: Rock Type of soil: Rock
Fault distance: 11 Fault distance: 22
f 1.12 f 1.17
s 0.13 s 0.057
SAmax(m/s2) 3.83 SAmax(m/s2) 1.90

Earthaquake record 990 Earthaquake record 7480

Seismical data Seismical data
Earthquake: Lazio Abruzzo, aftershock (Southern Italy) Earthquake: St. Die (France)
Date: 5/11/1984 Date: 2/22/2003
Magnitude: 5.5 Mw Magnitude: 4.7 Mw
Record data Record data
Station: Atina–Pretura Terrazza (Italy) Station: Bremgarten (Germany)
Type of soil: Rock Type of soil: Rock
Fault distance: 13 Fault distance: –
f 1.22 f 1.25
s 0.011 s 0.058
SAmax(m/s2) 6.11 SAmax(m/s2) 8.45

to calculate the seismic hazard. However, in this paper the

parametric probabilistic method has been used to present a
general method for the Iberian Peninsula. This method is appro-
priate for Spain and Portugal as the seismicity is diffuse and
moderate. However, for any location with enough information
available, other methods could be used. On a first attempt, the
authors tried to use specific attenuation laws for the Iberian
Peninsula. However, there were not enough records to carry out a
consistent regression analysis. So, the European Strong Motion
Database was used for the attenuation laws.
To select accelerograms the time of exposure of the structure,
the type of soil at the location and the admitted exceedance
probability must be established in order to obtain the seismic
hazard curves. With the hazard curves the USHARS can be plotted
and the design accelerograms can be selected.
 A. Morales-Esteban et al. / Soil Dynamics and Earthquake Engineering 43 (2012) 174–185
An example of the method proposed has been conducted for
San Pedro Cliff. A total of 10 design accelerograms have been
selected. It can be observed that the standard deviation for the
selected accelerograms is very low (a minimum of 0.011 for
record 990 and minor to 0.06 for 5 of the 10 accelerograms)
showing thus the robustness of the method.
Acknowledgements
The financial support given by the Spanish Ministry of Science
and Technology, projects BIA-2004-01302, BIA 2010-20377 and
TIN-68084-C02 and by the Junta the Andalucı́a, project P07-TIV-
02611 are acknowledged. The authors also want to acknowledge
Dr. Antonio Jesús Martı́n for providing the Spanish database of
earthquakes.
References
[1] Ambraseys NN, Douglas J, Sarma SK, Smit PM. Equations for the estimation of
strong ground motions from shallow crustal earthquakes using data from
Europe and the Middle East: horizontal peak ground acceleration and
spectral acceleration. Bulletin of Earthquake Engineering 2005;3:1–53.
[2] Andeweg B, de Vicente G, Cloetingh S, Giner J, Muñoz-Martı́n A. Local stress
field and intraplate deformation of Iberia: variations in spatial and temporal
interplay of regional stress sources. Tectonophysics 1999;305:153–64.
[3] Argus DF, Gordon RG, DeMets C, Stein S. Closure of the Africa–Eurasia–North
America plate motion circuit and tectonics of the Gloria fault. Journal of
Geophysical Research 1989;94:5585–602.
[4] Basili R, Valensise G, Vannoli P, Burrato P, Fracassi U, Mariano S, et al. The
Database of Individual Seismogenic Sources (DISS), version 3: summarizing
20 years of research on Italy’s earthquake geology. Tectonophysics
2008;453(1–4):20–43.
[5] Bernreuter DL, Savy JB, Mensing RW, Chen JC. Seismical hazard characteriza-
tion of 69 nuclear plant sites east of the Rocky Mountains. Technical Report
NUREG/CR-5250. Lawrence Livermore National Laboratory. U.S. Nuclear
Regulatory Commission; 1989.
[6] Bolt BA. Earthquakes and geological discovery. New York: Scientific American
Library; 1993.
[7] Bommer JJ, Douglas J, Strasser FO. Style-of-faulting in ground-motion
prediction equations. Bulletin of Earthquake Engineering 2003;1(2):171–203.
[8] Boore DW, Joyner WB, Fumal T. Estimation of response spectra and peak
acceleration from western North American earthquakes: an interim report.
Technical Report 93-509. U.S. Geological Survey Open-File; 1993.
[9] Boore DW, Joyner WB, Fumal T. The 2000 NEHRP recommended provisions
for seismic regulations for new buildings and other structures (FEMA 368)
and the accompanying commentary (FEMA 369). Technical Report 446.
Building Seismic Safety Council (BSSC); 2000.
[10] Buforn E, Udı́as A, Colombas A. Seismicity, source mechanism and tectonics
of the Azores–Gibraltar plate boundary. Tectonophysics 1988;152:89–118.
[11] Buforn E, Sanz de Galdeano C, Udı́as A. Seismotectonics of the Ibero–
Maghrebian region. Tectonophysics 1995;248:247–61.
[12] Calvert A, Sandvol E, Seber D, Barazangi M, Roecker S, Mourabit T, et al.
Geodynamic evolution of the lithosphere and upper mantle beneath the
Alboran region of the western Mediterranean: constraints from travel
tomography. Journal of Geophysical Research 2000;105:10871–98.
[13] Chandler AM, Cham LS, Lam NTK. Deterministic seismic hazard parameters
and seismic risk implications for the Hong Kong region. Journal of Asian Earth
Sciences 2001;20:59–72.
[14] Cornell CS. Engineering seismic risk analysis. Bulletin of the Seismological
Society of America 1968;58:1583–606.
[15] European Earthquake Database. /http://www.isesd.cv.ic.ac.ukS.
[16] Douglas J. A critical reappraisal of some problems in engineering seismology
(doctoral dissertation). UK: University of London; 2001.
[17] Douglas J. An investigation of analysis of variance as a tool for exploring
regional differences in strong ground motions. Journal of Seismology
2004;8(4):485–96.
[18] Douglas J. Use of analysis of variance for the investigation of regional
dependence of strong ground motions. In: Proceedings of the world con-
ference on earthquake engineering. Paper no. 29; 2004b.
185
[19] Ebel JE, Kafka AL. A Monte Carlo approach to seismic hazard analysis. Bulletin
of the Seismological Society of America 1999;89:854–66.
[20] Garcı́a-Monge J, Beltrán F, Sánchez-Cabañero JG. Seismic margin assessment
of Spanish nuclear power plants: a perspective from industry and regulators.
In: Proceedings of the OECD-NEA workshop on the seismic re-evaluation of
all nuclear facilities; 2001. pp 26–27.
[21] Giardini D, Grünthal G, Shedlock K, Zhang P The GSHAP global seismic hazard
map. International Handbook of Earthquake and Engineering Seismology,
IASPEI; 2003.
[22] Gra cia E, Palla s R, Soto JI, Comas M, Moreno X, Masana E, et al. Active faulting
offshore SE Spain (Alboran Sea): implications for earthquake hazard assess-
ment in the Southern Iberian Margin. Earth and Planetary Science Letters
2006;241(3–4):734–49.
[23] Gumbel JE. Statistics of extremes. Columbia University Press; 1958.
[24] Housner GW. Spectrum intensities of strong motion earthquakes. In: Pro-
ceedings of the symposium on earthquake and blast effects on structures;
1975. pp 25–33.
[25] International Code Council (ICC). International Building Code. Technical
Report 631; 2000).
[26] Jiménez-Munt I, Fernández M, Torne M, Bird P. The transition from linear to
diffuse plate boundary in the Azores–Gibraltar region: results from a thin-
sheet model. Earth and Planet Science Letters 2001;192:175–89.
[27] Joyner W, Boore DM. Peak horizontal acceleration and velocity from strong-
motion records including records from the 1979 Imperial Valley, California,
earthquake. Bulletin of the Seismological Society of America 1981;71(6):
2011–38.
[28] Justo JL, Azañón JM, Azor A, Saura J, Durand P, Villalobos M, et al.
Neotectonics and slope stabilization at the Alhambra, Granada, Spain.
Engineering Geology 2008;100:101–19.
[29] Kanamori H. The energy release in great earthquakes. Journal of Geophysical
Research 2001;80(20):2981–7.
[30] Kayabali K, Akin M. Seismic hazard map of Turkey using the deterministic
approach. Engineering Geology 2003;69(1–2):127–37.
[31] Kiratzi AA, Papazachos CB. Active crustal deformation from the Azores triple
junction to the Middle East. Tectonophysics 1995;243:1–24.
[32] Kochkin VG, Crandell JH. New Madrid seismic zone: overview of earthquake
hazard and magnitude assessment based on fragility of historic structures.
Partnership for advancing technology in housing. Technical Report 110, PATH
Research Center, Upper Marlboro; 2003.
[33] Konno T. Present and future seismic safety guideline for Npps in Japan.
Transactions SMiRT 2001;16:1117.
[34] Lin JP, Mahin SA. Effect of inelastic behavior on the analysis and design of
earthquake resistant structures. Earthquake Engineering Research Center.
Berkeley: University of California; 1985.
[35] Marcellini A, Daminelli R, Franceschina G, Pagani M. Regional and local
seismic hazard assessment. Soil Dynamics and Earthquake Engineering
2001;21:415–29.
[36] McGuire RK. Deterministic versus probabilistic earthquake hazards and risks.
Soil Dynamics and Earthquake Engineering 2001;21:377–84.
[37] Morales-Esteban A. Seismic hazard. Attenuation laws and earthquake time
series analysis (doctoral dissertation, Spanish). University of Seville /http://
fondosdigitales.us.es/media/thesis/1286/Q_Tesis_AME.pdf, 2010S.
[38] Morales-Esteban A, Martı́nez-Álvarez F, Troncoso A, de Justo JL, Rubio-
Escudero C. Pattern recognition to forecast seismic time series. Expert
Systems with Applications 2010;37(12):8333–42.
[39] Powell GH, Allahabadi R. Seismic damage prediction by deterministic
methods: concepts and procedures. Earthquake Engineering and Structural
Dynamics 2008;16(5):719–34.
[40] Romeo R, Prestininzi A. Probabilistic versus deterministic hazard analysis: an
integrated approach for sitting problems. Soil Dynamics and Earthquake
Engineering 2000;20:75–84.
[41] Romeo R, Pugliese A. Seismicity, seismotectonics and seismic hazard of Italy.
Enginering Geology 2000;55:241–66.
[42] Trifunac MD, Lee VM, Anderson JG. Methods for introduction of geologic data
into characterization of active faults and seismicity and upgrading of the URS
technique. Technical Report NUREG/CR, vol. 2. Nuclear regulatory commis-
sion; 1987.
[43] UNDRO. Natural disasters and vulnerability analysis. Technical report. Report
Expert Group Meeting, Geneva; 1979.
[44] Vanucci G, Gasperini P. The new release of the database of Earthquake
Mechanisms of the Mediterranean Area (EMMA Version 2). Annals of
Geophysics 2009;47:307–34.
[45] Veneziano D, Cornell CA, O’Hara T. Historical methods of seismic hazard
analysis. Technical Report NP-3438. Electric Power Research Institute, Palo
Alto, California; 1984.