MARMARAS EARTHQUAKE REHABILITATION
PROJECT
An Integrated Intervention for the Rehabilitation of Duzce
Earthquake Hazard Assessment
The case of DUZCE
31/ 04/ 2004
Report No. 1
Contents
1
Introduction ............................................................................................................. 3
2
Location and Topography ...................................................................................... 4
3
Regional seismo- tectonic setting ........................................................................... 5
4
Geotechnical zonation of the Duzce municipal area .......................................... 13
5
Deterministic seismic hazard evaluation for the city of Duzce ......................... 14
5.1
Reference earthquake ...................................................................................... 14
5.2
Attenuation of ground motions ....................................................................... 15
5.3
Results of deterministic approach ................................................................... 16
5.4
Comparison with other studies........................................................................ 18
6 Probabilistic seismic hazard evaluation for the city of DUZCE ....................... 19
6.1
Computational tools ........................................................................................ 19
6.2
Description of the earthquake catalogue ......................................................... 19
6.3
Seismotectonic zonation criteria ..................................................................... 19
6.4
Final PGA zoning for different periods .......................................................... 21
7 Conclusions ............................................................................................................ 22
8
References .............................................................................................................. 25
1 Introduction
This study “Earthquake hazard Assessment” was an attempt to estimate the seismic
hazard for the city of Duzce. The main goal of that report is to give very briefly the
steps and the uncertainties derived from the performing a probabilistic and a
deterministic seismic hazard study for the city of Duzce. This study uses all available
sources of information, technical papers and research works. The seismotectonic
environment of the area is very active although there are several uncertainties (until
now) in the characteristic of the seismic source zones, the fault length, the magnitudes
assigned to the faults, the seismicity parameters, the seismic history of the area, the
attenuation relationships etc. General, the level of knowledge of the area, the specific
geology and the paleoseismic information is quite low as several research should be
conducted to provide more reliable seismological information of the area.
The ground motion scenario that was developed for this study is:
Deterministic scenario: One deterministic scenario was developed. Because the
seismic catalogue of historical earthquakes have very large uncertainties (the
epicentres maybe differ even 10km) as it is not based on good paleoseismic data and
very long and reliable historical data, only a maximum credible earthquake was
chosen that is Duzce earthquake (1999)
Maps are given in GIS format.
Probabilistic scenarios: one probabilistic scenario was developed: for the dominant
SSZ that corresponds to fault cases. The selected software for the probabilistic seismic
hazard assessment is CRISIS 99 (Ver. 1.018) developed by M. Ordaz and co- workers
of the Institudo de Ingenieria de la U.N.A.M. In CRISIS 99, the sources are modelled
as area sources (Seismic Source Zones or SSZs), fault sources or point sources.
Seismic hazard in Duzce was performed using uniform seismic zones (SSZ):
The followings assumptions were made:
- The selection of the zones and the seismicity parameters a and b for each zone that
affect Duzce is based on the researches of Kayabali k, 2002.
- The standard deviation of the maximum magnitude for the earthquakes occurred
between 1900- 2000 is 0.25. A threshold magnitude M=4.0 is selected.
- For the attenuation relationships, Ambraseys et al. (1996) relationship was
assigned in order to perform a probabilistic study. Three other local attenuation
relationships (Inan et al ,1996 ; Aydan et al, 1996; Ozbey, 2000) were used for
deterministic analysis in order to compare the results with the real records of
Duzce earthquake
- Seismic zones were supposed to be plane, parallel to surface.
- Focal depth was selected equal to 10km for the case of SSZ as the area is
characterized by shallow seismicity.
- The grid used for the analysis for SSZ case is 1000*1000m.
The following maps, in GIS format, will be presented for 100, 475 years recurrence
interval.
2 Location and Topography
Duzce is just situated between Ankara and Istanbul; Ankara is 240km away to the
East and Istanbul is 228km away to the West. The road of D-100 passes through
Duzce and TEM Highway passes around it. Duzce is placed into the plateau of The
West Blacksea coast. The city is surrounded to the West by Sakarya, to the Northeast
by Zonguldak and to the East by Bolu. The distance from East to West is 23 km and
from North to South is 20 kms. It is a very well- known province with an area size is
1.014km2.
Figure 1 Location of Duzce
Figure 2 Topography of Marmara sea region. Altitudes ranges from –1200m (blue) to
+2500m (red) (Gurbuz G et al, 2000)
Figure 3 Gradient of topography in N30 direction (Gurbuz G et al, 2000)
3 Regional seismo- tectonic setting
The majority of earthquakes happened to Turkey can be attributed to the relative
movements of Eurasian Plate, African Plate and Arabian Plate which is still in
progress. The Arabian/African and Eurasian plates move north and south towards
each other with a result, Turkey is being squeezed out westwards.
Figure 4 World –wide seismogenic plates
There are three main sources of seismic activity in Turkey, which are:
North Anatolian Fault (NAF): The North Anatolian Fault is a morphologically
distinct and seismically right-lateral strike-slip fault. It has a well-developed surface
expression for most of its length of 1300km. North Anatolian Fault system is one of
the most seismically active right-lateral strike-slip faults in the world. Since 1939
there have been 9, M=7.0 or larger earthquakes along the fault. These earthquakes
followed a systematic pattern that progressed generally from east to west along the
fault system.
East Anatolian Fault (EAF): The East Anatolian Fault is an active left-lateral strikeslip fault, which extends from Antakya to Karliova, the eastern terminal of NAF. It is
a fault zone that is about 2-3 km wide, and links into the Dead Sea Fault System.
Western Turkey Graben Complex: This is an area of intense seismic activity that is
related to the east-west trending graben complexes in the Aegean region.
Figure 5 Aerophoto of Marmara region and location of NAF
Generally, the Anatolian plate is characterized by shallow earthquakes (<30km). The
latest can be noticed from Figure 6.
Figure 6 Shallow seismicity in Anatolian plate (all magnitude)
Duzce is located nearby the North Anatolian Fault (NAF) and next to Duzce fault that
is small branch of NAF. Earthquakes on the North Anatolian fault are caused by the
northwards motion of the Arabian plate against the Eurasian plate, squeezing the
small Turkish microplate westwards. Also, compression in this region is due to the
northwards motion of the African plate, which produces subduction at the Cyprus and
Hellenic arcs. The small Turkish microplate is bounded on the east by the East
Anatolian fault zone (EAFZ), on the north by the North Anatolian fault zone (NAFZ),
on the west by a diffuse zone of deformation surrounding the greater Aegean region,
and on the south by the Hellenic and Cyprus arcs.
The relative movement of Eurasian Plate and Anatolian plate provoke ten major
earthquakes (Ms>7.0) in the last century:
- Erzincan earthquake (26-12-1939), MS=7.9- 8.0, fault length about 360km, 30.000
deaths.
- Erbaa earthquake (20-12-1942), MS=7.1, fault length about 50km.
- Tosya earthquake (26-11-1943), MS =7.6, fault length about 280km.
- Bolu-Gerede earthquake (01-02-1944), MS=7.3, fault length about 165km.
- Abant earthquake (26-05-1957), MS=7, fault length about 30 km.
- Mudurnu Valley earthquake (22-07-1967), MS=7.1, fault length about 80 km.
- Kocaeli earthquake (17-08-1999), MS=7.8,
- Duzce earthquake (12-11-99), MS=7.2, 45km surface rupture, 1.000 deaths
Especially the epicenters of the more recent ones (Kocaeli, Duzce) was on NAF and
these can be noticed clearly in Figure 7.
Figure 7 The epicenters of Kocaelli and Duzce earthquake
Although this region is one of the more seismic active areas in world- wide there was
no reliable recordings or important references from previous earthquakes. More
reliable data are given for the last century and especially in the last two decades.
More representative of the seismicity on NAF are the figure following:
Figure 8 Seismic Activity in Turkey since 1939
However, geologic information on the past activity of the fault has been very limited
until late 1980s when detailed survey of recent ruptures including exploratory
trenches is introduced to Turkey. At the same time, historic documents were so sparse
that only a few events of 1668 A.D. (Ambraseys and Finkel, 1988) and of 1784 A.D.
(Barka, 1992) and a few periods of higher activity (Ambraseys, 1975) have been
inferred. Turkey-Japan joint research on the paleoseismology of the North Anatolian
fault began in 1989 and has investigated three segments of 1939, 1944, and 1784?
[There has been no surface faulting in this century according to Barka (1992). Now
for the first time, it is being possible to compare time-series of surface faulting events
on different segments. In a trench east of Gerede, on the 1944 (Bolu-Gerede) surface
fault, we recognized 8 earthquake events since around 30 B.C. Average recurrence
interval is estimated to be between 200 and 300 years. On the 1784 surface fault, a
trench east of Erzincan showed 5 events during these about 1000 years.
A diagram of the earthquakes occurred in last century on Marmara area, is presenting
in Figure 9.
Figure 9 Seismicity of Turkey between 1900 and April 2003 (data obtained from the
Earthquake Research Center of the General Directorate of Disaster Affairs of Turkey)
More detailed list of the earthquakes that occurred in the last century is given from
Alsan (1976) while a more descriptive list for the millennium (without the epicenter
of the earthquakes) is given by Sahin M, Tari E, 2000.
Table 1 List of earthquakes in the last century (Alsan, 1976)
Table 2 Damaging earthquakes in the eastern Marmara region that occurred during the
last millennium (Sahin M, Tari E, 2000)
Because they’re several uncertainties in the epicenter of the earthquakes a range of
zone assigned to earthquakes (Ambraseys & Finkel, 1987; Ambraseys & Finkel,
1995; Barka, 1992). Zones for earthquakes before 1939 are based largely on
isoseismals and are thus approximate. The historical seismicity is presented in Figure
10.
Figure 10 Historical seismicity on North Anatolian Fault
A more complicated form of the faults is provided by Yaltirak et al (1998) beside with
table that gives maximum magnitude corresponds to the active faults.
Figure 11 Map of tectonic elements of Turkey compiled by Yaltirak et al (1998)
Taking into the assumption that 1/3 to ½ of the total length of fault would rupture
when it generates the maximum earthquake (Mark, 1977) and Wells and Coppersmith
(1994) proposed relationships between fault rupture length at surface and the
earthquake magnitude for strike slips, maximum magnitude was assigned (Table 3) .
Mw= 5.16 + 1.12logL
Table 3 Active faults and corresponding maximum magnitudes
Fault
North Anatolian
Duzce
Length
(km)
1400
60
Fault type
Reference
SS
SS
Saroglu et al (1992)
Saroglu et al (1992)
Maximum
magnitude (Mw)
8.1
6.5
4 Geotechnical zonation of the Duzce municipal area
A detailed geotechnical zonation of Duzce is based on the physical, mechanical and
dynamic properties of soil formations. A simplified approach in order to perform a
seismic hazard study was made using Ambrasey’s (1996) classification. According to
Ambraseys (1996) the classes of site geology are defined by the following ranges of
average Vs velocities over the upper 30m of the site: Vs30: Rock (R)>750m/sec, Stiff
Soil (A) 360- 750m/sec, Soft Soil (S) 180- 360m/sec and very Soft Soil (L)<
180m/sec. As records for very soft soil are very restricted, soft & very soft soil
categories grouped into one category called “Soft Soil”.
Table 4 Classification of soil conditions according to Ambrasey’s (1996) classification
Classification
Vs,30 (m/ sec)
Rock
> 750
Stiff Soil
360- 750
Soft Soil
< 360
Figure 12 A simplified Geotechnical map for Duzce using Ambrasey’s classification
5 Deterministic seismic hazard evaluation for the city of Duzce
Duzce earthquake that occurred in 12/11/1999 is selected as a representative
earthquake of an area. It is the more recent one, with well- recordings and with major
damages in the city. Comparing the magnitudes and the distances (as they can be
calculated between the epicenter and the Meteology station) of the more destructive
earthquakes in the last century, Duzce (1999) earthquake is the more severe one.
5.1 Reference earthquake
Duzce earthquake (11-12-1999): Epicenter (31.15E, 40.77N), shallow earthquake
(depth: 10.0km), Ms= 7.2
On November 12, 1999 at 6:57 PM (local time), a magnitude Mw 7.2 earthquake
struck the Dόzce-Bolu area of Turkey, 70 km east of Adapazari and 170 km northwest
of Ankara. The earthquake epicenter is located near the town of Dόzce, on the eastern
end of the fault that is believed to have ruptured during the August 17, 1999, Mw =
7.4 event. The November 12, 1999 earthquake epicenter is located at 40.77N 31.15E,
which is about 110 km east of the magnitude 7.4 main shock on August 17 which
killed over 17,000 people and injured another 50,000. On November 26, 1999,
preliminary estimates of casualties and injured are 755 and 4948, respectively. The
material damage is extensive. 1342 structures have collapsed. A total of 7081
residential and industrial buildings are heavily damaged. Table 5 summarizes the
distribution of losses due to the November 12, 1999 earthquake to Duzce. Most of the
damage is concentrated in Kaynasli a small town on the main highway between Duzce
and Bolu. Loss of life in Duzce seems to concentrate in few collapsed buildings that
were "lightly" damaged in the August 17 earthquake, superficially repaired and later
inhabited. Fire damage due to overturning of coal and wood burning stoves and
explosion of bottled LNG. The main highway system between Istanbul Ankara
closely follows the Duzce Fault and the whole area is mountainous and highly prone
to landslide hazard. At Bakacak section (between Kaynasli and Bolu) of the main
highway between Istanbul and Ankara two lanes of the four-lane highway collapsed
due to land slide for a section of about 200m. Fortunately no vehicles were on the
highway at the moment of collapse. The major viaducts in the regions are structurally
intact except possible damaged to the electrometric bearings under the decks. The
Bolu highway tunnel is under construction near the ruptured area.
Table 5 Distribution of losses in Duzce as a result of Duzce earthquake
(A. Ansal et al ,1999)
Location
Duzce
Number of
Casualties
Number of
Injured
344
2,800
Medium Structural Damage/
Industrial
49
Structures
to be Torn
Down
Heavy
Structural
Damage/
Residential
617
3,588
Light Structural Damage/
Residential
174
Heavy
Medium
Structural
Structural
Damage/
Damage/
Industrial
Residential
874
152
Light Structural Damage/
Industrial
47
Figure 13 The location of epicenter of Duzce earthquake
on Duzce fault (branch of NFA)Ghasemi H et al (2000)
5.2
Figure 14 Location of
epicenter of Duzce in
correlation with Ankara
Attenuation of ground motions
The relationship that was selected in order to propose the earthquake hazard
assessment (Deterministic/ probabilistic) is Ambrasey’s 96. A brief description is
given below.
Ambraseys et al (1996)
AMB96 (Ambraseys et al., 1996) provides acceleration response spectral values.
Ambrasey’s had classified the soil into 4 classes (from rock to very soft soil) almost
coincident with the Eurocode8 soil classification (CEN, 2001). Ambraseys was based
on a European database that includes MS ranging between 4.0 and 7.5.
Ambraseys et al. (1996) attenuation relationship is the following:
log (y)= C1΄+C2*M+ C4*log (r) +Ca*SA+ Cs*Ss+ σ*P
Given that: r=
d 2  h02
The parameter y represents peak horizontal ground acceleration in g (related to
structural period) while d is the shortest distance from the station to the surface
projection of the fault rupture (in km). The parameter ho is a constant and alternate
according to structural period, as well as C1΄, C2, C4, Ca, Cs (Table 6) for a given
structural period).The standard deviation of log(y) is σ and the constant P takes a
value of 0 for mean values and 1 for 84- percentile values of log(y). In our case, we
considered that P=0.
The parameters Sa and Ss take the following values:
Rock: Sa=0, Ss= 0
Stiff Soil: Sa= 1, Ss=0
Soft Soil: Sa= 0, Ss=1
Table 6 Constants given according to structural period (Ambrasey’s 96)
T
0
0.1
0.15
0.2
0.28
0.32
0.4
0.5
0.6
0.7
0.8
0.9
1
1.4
1.6
C1'
-1.48
-0.84
-0.98
-1.21
-1.46
-1.63
-1.94
-2.25
-2.49
-2.67
-2.86
-3.03
-3.17
-3.52
-3.68
C2
0.266
0.219
0.247
0.284
0.326
0.349
0.377
0.42
0.438
0.463
0.485
0.502
0.508
0.522
0.52
h0
3.5
4.5
4.7
4.2
4.4
4.2
3.6
3.3
2.5
3.1
3.7
4
4.3
3.4
2.5
C4
-0.922
-0.954
-0.938
-0.922
-0.946
-0.932
-0.888
-0.913
-0.881
-0.914
-0.925
-0.92
-0.885
-0.839
-0.781
Ca
0.117
0.078
0.143
0.135
0.134
0.125
0.139
0.147
0.124
0.116
0.127
0.124
0.128
0.109
0.108
Cs
sigma ln(10^sigma)
0.124 0.25
0.576
0.027 0.27
0.622
0.085 0.27
0.622
0.142 0.27
0.622
0.158 0.29
0.668
0.161 0.31
0.714
0.172 0.31
0.714
0.201 0.32
0.737
0.212 0.32
0.737
0.214 0.33
0.760
0.218 0.32
0.737
0.225 0.32
0.737
0.219 0.32
0.737
0.197 0.31
0.714
0.206 0.31
0.714
The Ambrasey’s relationship giving PGA peak horizontal ground acceleration with
respect to 416 records for which site conditions have been classified is the following:
log(a)= -1.48+0.266*Ms- 0.922*log (r)+0.117Sa+0.124Ss+0.25P
where: ho= 3.5 and Sa, Ss, P were the same given in Ambrasey’s attenuation former
described.
5.3
Results of deterministic approach
A comparison between the results obtained by Ambrasey’s relationship and three
local attenuation relations for Turkey Inan et al (1996), Aydan et al (1996) and Ozbey
(2000) is given for Meteorology station.
Table 7 Turkish Attenuation relationships
Relationship by
Inan et al (1996)
Aydan et al (1996)
Ozbey (2000)
Empirical form
log PGA (gal) =0.65M-0.9log(r) -0.44
PGA (gal)=2.8*(e0.9M * e-0.025r-1)
log PGA (grams)=-2.6517+0.4524Mw-0.986log(r2+h2)0.5
Where:
M, magnitude (=Ms when greater than 6.5; ML otherwise)
Mw, Moment magnitude
r, epicentral distance from fault (km)
h=7
Ozbey (2000) relationship was developed using data from Kocaelli and Duzce
earthquake. The relationship is valid for strike-slip faulting, soft or stiff soil sites and
for Mw between 4.0 and 7.4. This relation was found in Durukan E (2002) paper.
Table 8 Comparison of the results PGA obtained by 4 attenuation relationships in
Meteorology station
Location
Meteorology
station
(31.17, 40.85)
Ambrasey’s (1996)
Inan et al (1996)
Aydan et al (1996)
Ozbey (2000)
0.44g
2.18g
1.35g
1.43g
As it can be noticed the results have very large scattering. The results obtained using
Ozbey (2000) and Aydan et (1996) relationship are quite close but quite far from the
Inan et (1996) and Ambrasey’s 96. The recorded acceleration in Meteorology station
is 0.55g which is quite near to the acceleration calculated by Ambrasey’s 96
relationship while it is quite different from the results obtained by Aydan et al. It can
be seen that maybe Ambrasey’s 1996 underestimate the expected PGA in
Meteorology station. Ozbey relationship is the same for soft and stiff soil sites as it
depends only upon the magnitude and the distance and derived only from the data of
Duzce and Kocaeli earthquake. This relationship hasn’t been tested in other
earthquakes in Turkey while Ambrasey’s 96 has been tested world-wide, it uses data
from Turkey region and it is neared to real records. That is the main reason the
Ambrasey’s relationship was selected to conduct this study. Quite large accelerations
(two time above the recorded) are expected by using Inan et relationship.
A PGA map spatial distributed in Duzce is provided (Figure 15) using Ambrasey’s
96-attenuation relationship.
Figure 15 PGA values using Duzce earthquake (1999)
5.4
Comparison with other studies
In this study, Duzce earthquake was selected as the appropriate deterministic scenario
for Duzce city and was compared with the results of Kayabali K et al, 2003
deterministic approach for all Turkey. Although this comparison is quite indicative, as
it has nothing similar in the approach used for Duzce deterministic scenario although
is very interesting to see that the results aren’t differ greatly from the calculated in this
study. The main difference of isoaccelation map for all Turkey (Figure 16) presented
by Kayabali K et al (2003) is that was derived using TUMDES code after the
identification of all faults, calculation of the closest distance, assignment of a
magnitude and providing an attenuation relationship to each fault. Moreover, as it
wasn’t possible to know the specific soil conditions of all Turkey, the attenuation
relationship that it was assigned in each fault was corresponded to rock sites.
Expect all the above differences a comparison was conducted for a specific town
(Duzce) and more for a specific point in Duzce (Meteorology station) with a specific
topography, geological condition and local soil conditions (soft soil)- see Table 9.
Figure 16 Iso-acceleration map based on the relationship by Sadigh et al (1997)
produced by Kayabali K et al (2003)
Table 9 Comparison of PGA between Kayabali K (2003) study and this study
Duzce
Kayabali K et al (2003)
PGA
0.2g (rock)
This study (Ambrasey’s 1996)
0.44g (soft soil)
if it was rock: 0.33g
It can be noticed that although the great uncertainties and the different way of
approaching deterministic scenario no large scattering of the results was assigned.
The reason of performing comparisons as the ones in Table 8 or Table 9 is to check
the results that we obtained by making some specific assumption with other results
that are derived using different attenuation or even different method as the
uncertainties are very important.
6 Probabilistic seismic hazard evaluation for the city of DUZCE
The probabilistic seismic analysis for the city of DUZCE is performed taking into
account SSZ zones that responds to fault cases. The seismic hazard was performed
using Poissonian approach and the return period that was calculated was 100 and 475
years.
6.1 Computational tools
The Probabilistic Seismic Hazard Analysis (PSHA) for the city of Thessaloniki was
conducted using CRISIS 99; a code- program developed by Professor M. Ordaz and
co- workers of the “Institudo de Ingenieria de la U.N.A.M”, Mexico City. In CRISIS
99, the sources are modelled as area sources (Seismic Source Zones or SSZs), fault
sources or point source. The present application is performed based on area geometry
modeling for sources 1, 4.
6.2 Description of the earthquake catalogue
Although previously, an earthquake catalogue was given (Table 1, Table 2) the
epicenter distribution of earthquakes in the general region of Turkey occurred earlier
than the 1970s do not correlate very well with the neotectonic elements which
attributed mostly to the quality of instrumental records as well as the calculation
techniques employed. The present- day calculations of epicenters of those earthquakes
show deviations as much as several tens of kilometers from the ones determined
through earlier computation. On the other hand, the more accurate but relatively short
term data for the last three decades do not provide a satisfactory correlation between
the epicenter distributions and neotectonic features (Kayabali K, 2002).
6.3 Seismotectonic zonation criteria
The seismic sources selected to perform a seismic hazard analysis were determined by
Kayabali K (2002). A total of 14 seismic sources were delineated in Turkey based
mostly on Erdik et al (1985) and Yaltirak et al (1998) works.
Duzce is influenced by seismic source zone 1 and 2 (Figure 17) with regression
coefficient, fault lengths and maximum magnitudes for seismic zones given in Table
10.
Table 10 Seismicity parameters for seismogenic sources (Kayabali k, 2002)
Source
zone
1
2
Erdik et al
(1985)
a
b
4.1
0.78
3.6
0.8
Alptekin1
(1978)
a
b
5.66 0.71
4.41 0.55
(1): Smallest squares
(2): Generalized smallest squares
(3): Maximum likelihood
Alptekin2
(1978)
a
b
4.86 0.55
4.27 0.52
Alptekin3
(1978)
a
b
3.95 0.45
Kayabali K
(2002)
a
b
4.9 0.67
4.0 0.63
Fault length
(km)
Mmax
1500
1500
8.2
8.2
Figure 17 Seismic sources used in the probabilistic analysis (Kayabali K, 2002)
The assumption for the seismic source regionalization with the study of Kayabali K
(2002) is that earthquakes only occur on fault. Although most seismic zonation maps
allow for existence areas of seismic zones, the uncertainty with the distribution of
epicenters of past earthquakes (even 10km) crosses out this opinion.
The Rate λ (number of earthquakes greater than the magnitude Mo =4 was calculated
by the equation (Richter, 1958)
Log λ= a + b*M
Where: a, b the regression coefficient given in Table 10
Vertices of the seismic zones
Seismic Zone 2
Long. (o)
Lat. (o)
30
39.33
34.5
39.81
39.33
39
39.66
39.50
34.5
40.11
30
39.66
Seismic Zone 1
Long. (o)
Lat. (o)
22.67
38.68
22.67
38.14
30
39.33
30
39.66
27.5
40.07
The magnitude scale that is not subject to saturation is the moment magnitude since it
is based on the seismic moment, which is a direct measure of the factors that produce
rupture along the fault. The magnitude that was given as an input in CRISIS 99 is
moment magnitude (Mw), which for 5.5 Ms 7.2 is equal to Ms (Ms  Mw), while
the standard deviation of M=0,25 (as a result of several uncertainties).
The seismic hazard model that it was used is Poisson model. The basic assumption is
that seismic events are spatially and temporally independent and the probability that
two seismic events that will take place at the same location and at the same time
approaches zero.
The Poisson model is described by the following equation:
Pr(N=n/ ν,t) = (e-νt (vt)n)/n!
Where: Pr is the probability of n events during time period t with the mean rate of
occurrence, ν.
6.4 Final PGA zoning for different periods
The grid used to perform the analyses with CRISIS 99 was 1000*1000m. The grid has
origin (31.13, 40.82) and extents 14km in longitude and in latitude direction in order
to include the new city of Duzce.
Figure 18 Grid selected in correlation with seismic zones
Two maps for PGA where obtained for 100years and for 475years using the abilities
of ArcGIS 8.2.
PGA values (in gal) for T=100y
410
402
395
388
381
373
366
359
Figure 19 PGA values for the city of Duzce (100years recurrent period)
PGA values (in gal) for T=475y
928
912
896
880
864
848
832
816
Figure 20 PGA values for the city of Duzce
(475years earthquake recurrent period)
7 Conclusions
A comparison between the results of PSA (spectral accelation) derived from the
deterministic and probabilistic approach in a soft soil site (Meteology station) using
Ambrasey’s relation (1996) and the real record (Dimitra) is given in Figure 21.
Comparison of PSA (gal) in Meteorology station -Duzce
10000
100year
PSA (gal)
9000
475year
8000
1000year
7000
2000year
6000
50year
5000
Deterministic scenario
4000
3000
2000
1000
0
0
0.5
1
1.5
2
Period (sec)
Figure 21 Comparison of PSA values in Meteorology station
(deterministic, probabilistic approach for 50, 100, 475, 1000, 2000years
and the real record)
Table 11 Comparison of PGA (g) values in Meteorology station
P
G
A
(g)
Real record
(longitude)
Real record
(transversal)
Deterministic
scenario
50y
100y
475y
0.55
0.38
0.44
0.25
0.39
0.89
1000y 2000y
1.29
It should be noticed (Table 11) that the real record in the longitude direction in the
specific site was 0.55g, between 100 and 475years based on probabilistic approach.
Also, a comparison of the probabilistic seismic hazard obtained by Crisis 99 and by
the map obtained by Kayabali K (2002) using SISTEHAN-II code (Poisson model,
Joyner & Boore, 1988 attenuation relationship) is provided. Although the results
obtained by Kayabali K (2002) for the Turkey region and especially for Duzce (0.1g
for 100years, 0.2g for 475years) is quite low comparing with ones obtained by
CRISIS 99 and the one derived from real records.
1.82
Figure 22 Isoacceleration contour maps for the return period of 100years (in g)
Figure 23 Isoacceleration contour maps for the return period of 475years (in g)
If we assume that Duzce earthquake is an event that has a recurrence period about
220years and the records during Duzce earthquake shown a PGAmax=0.55g the results
obtained by the above analysis is very optimistic. The results given by Kayabali K are
referred to all Turkey area and for rock conditions while in Duzce there are soft and
stiff soil sites mainly. Table 12 shows the differences in terms of PGA between values
calculated by CRISIS 99 (100, 475years), Kayabali K study (2002) and the real record
(Duzce earthquake, 1999).
Table 12 Difference in PGA values (probabilistic approach)
Location
Duzce
CRISIS 99
100year 475years
0.394g
0.894g
Kayabali K (2002) study
100year
475years
0.1g
0.2g
Real record
Duzce earthquake
0.55g
8 References
Alsan E, Tezucan L, Bath M (1976) “An earthquake catalogue for Turkey for the
interval 1913-1970”, Tectonophysics 31, pp.13-20.
Ambraseys, N. N., Finkel, C. F. (1987) “Seismicity of Turkey and neighboring
regions, 1899-1915”. Annales Geophysicae 5, pp. 701-725.
Ambraseys, N. N., Finkel, C. F. (1995) “The Seismicity of Turkey and Adjacent
Areas: A historical review, 1500-1800” Muhittin Salih EREN, Istanbul.
Ambraseys N.N. (1995) ”The prediction of earthquake peak ground acceleration in
Europe”. Publ. Earthquake Engineering and Structural Dynamics, 24, pp.467- 490.
Ambraseys N.N, Simpson K.A, Bommer J.J. (1996) “Prediction of horizontal
response spectra in Europe”. Publ. Earthquake Engineering and Structural Dynamics,
25, pp. 371- 400.
Ansal, J. P. Bardet, A. Barka, M. B. Baturay, M. Berilgen, J. Bray, O. Cetin, L. Cluff,
T. Durgunoglu, D. Erten, M. Erdik, I. M. Idriss, T. Karadayilar, A. Kaya, W. Lettis,
G. Olgun, W. Paige, E. Rathje, C. Roblee, J. Stewart, and D. Ural (1999) “Initial
Geotechnical Observations of the November 12, 1999, Dόzce Earthquake” A report of
the Turkey-US geotechnical earthquake engineering reconnaissance team.
Ayadan O, Sezaki M, Yarar R (1996) “The seismic characteristics of Turkish
earthquakes” 11th World Conference on Earthquake Engineering, Acapulco, Mexico,
CD paper no 1025
Barka A.A., (1992) ”The North Anatolian fault zone” Ann.Tectonicae, 6, pp.164-195.
Durukal E (2002) “Critical evaluation of strong motion in Kocaeli and Duzce
(Turkey) earthquakes” Soil Dynamics and Earthquake Engineering 22, pp.589-609.
Erdik M, Doyuran V, Gulkan P, Akkas N (1985) “A probabilistic assessment of the
seismic hazard in Turkey” Tectonophysics 117, pp. 295- 344
Ghasemi H, Cooper J, Imbsen R, Piskin H, Inal F, Tiras A (2000) “The November
1999 Duzce Earthquake: Post- Earthquake Investigation of the structures on the TE”
Publication No.FHWA-RD-00-146.
Inan E, Colakoglou Z, Koc N, Bayulke N, Coruh E (1996) “Earthquake catalogues
with acceleration records from 1976-1996” General Directorate of Disaster Affairs,
Earthquake Research Division, Ankara, Turkey, 98pp (in Turkish).
Joyner W.B, Boore D.M (1988) “Measurement, characterization, and predition of
strong ground motion” Earthquake Engineering and Soil Dynamics, 2, Recent
Advances Ground Motion Evaluation, pp.43-102
Kayabali K, Akin M (2002) “Modeling of seismic hazard for Turkey using the recent
neotectonic data” Engineering Geology 63, pp.221-232.
Kayabali K, Akin M (2003) “Seismic hazard map of Turkey using the deterministic
approach” Engineering Geology 69, pp.127-137.
Mark R.K (1977) “Application of linear statistical model of earthquake magnitude
versus fault length in estimating maximum expectable earthquakes” Geology 5,
pp.464-466.
Okumurea K., Yoshioka T, Ismail Kuscu (1993) “Surface Faulting on the North
Anatolian Fault in These two Millennia”, U.S. Geological Survey Open-file Report,
94-568, p 143-144.
Ozbey C (2000) “Empirical peak horizontal acceleration relationship for
northwestern Turkey”, MSc Thesis, Bogazici University, Istanbul.
Richter C.F (1958) “Elementary Seismology” Freeman, San Francisco CA, pp. 768
Sahin M & Tari E (2000) “The August 17 Kocaelli and the November 12 Duzce
earthquakes in Turkey” Earth Planets Space, 52, pp. 753-757
Saroglu F., Emre O, Kuscu I (1992) “Active fault map of Turkey“ Printed by General
Directorate of Mineral Research and Exploration.
Yaltirak C, Alpar B, Yuce H (1998) “Tectonic elements controlling the evolution of
the Gulf of Saros (Northeastern Aegean Sea, Turkey), Tectonophysics 300, pp.227248
Wells D.L Coppersmith K.J (1994) “New empirical relationships among magnitude,
rupture length, rupture width, rupture area and surface displacement” Bull. Seismol.
Soc. Am. 4 (84), pp.975-1002