Anda di halaman 1dari 27

Engineering Geology 74 (2004) 265 291 www.elsevier.

com/locate/enggeo

An attenuation relationship based on Turkish strong motion data and iso-acceleration map of Turkey
Resat Ulusay *, Ergu n Tuncay, Harun Sonmez, Candan Gokceoglu
Hacettepe University, Department of Geological Engineering, 06532 Beytepe, Ankara, Turkey Received 27 May 2003; accepted 13 April 2004 Available online 9 June 2004

Abstract This paper presents an attenuation relationship of peak ground acceleration (PGA) derived from Turkish strong motion data for rock, soil and soft soil sites and an iso-acceleration map of Turkey based on this relationship. For the purpose, among all the three-component accessible records, 221 records from 122 earthquakes that occurred in Turkey between 1976 and November 2003 were selected. The database was compiled for earthquakes with moment magnitudes (Mw) and PGA values ranging between 4.1 and 7.5, and 20 and 806 gal, and distances to epicenter considered in the database were between 5 and 100 km. From the regression analysis of the data, an attenuation equation of PGA considering rock, soil and soft soil conditions was developed. The PGA values predicted from the equation suggested in this study and those both from a few domestic equations and some imported equations were compared. In addition, an iso-acceleration map of Turkey was constructed using the suggested attenuation equation and considering both known active faults and epicenter locations of the earthquakes that have occurred in Turkey. D 2004 Elsevier B.V. All rights reserved.
Keywords: Attenuation relationship; Iso-acceleration map; Peak ground acceleration; Strong motion database; Turkey

1. Introduction In seismic hazard analyses the quantitative description of the ground motions are very important. One of the ground motion parameters commonly used in geotechnical and structural engineering analyses is peak ground acceleration (PGA). Therefore, estimation of this parameter in a precise manner has a prime

* Corresponding author. Fax: +90-312-299-2034. E-mail address: resat@hacettepe.edu.tr (R. Ulusay). 0013-7952/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.enggeo.2004.04.002

importance in engineering design. Major initiatives to instrument seismically active regions around the world were undertaken in the twentieth century, and these instruments have provided a large inventory of recordings. Data from this inventory are used to develop empirical strong motion attenuation relationships to estimate earthquake ground motions based on some characteristics of the earthquakes and local geology. PGA is the simplest strong-motion parameter and hence more than 120 attenuation equations have been derived in the past to predict it (Douglas, 2003). However, these equations were derived for different

266

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

earthquake regions and fault types, and interplate versus intraplate. In addition, their data selection criteria are different, and some of them pertain to only a single ground type such as rock or firm soil. As a result of their nature, differences among the estimated PGA values from the existing attenuation relationships from one region or country to another result in a limitation in their use. Therefore, the use of attenuation relationships derived from the records of a region, where the predictive equations are considered, shows an increasing tendency between the associated engineering community. In Turkey, the seismic hazard zonation map was published by the Ministry of Public Works and Settlement of Turkey (1996). Based on this map, Turkey is divided into five subclasses of seismic zone with PGA values of >0.4g, 0.3 0.4g, 0.2 0.3g, 0.1 0.2g and < 0.1g for zones ranging from I to V, respectively. The current practice in Turkey is to directly use the PGA values from the seismic codes published by the Ministry of Public Works and Settlement of Turkey (1998). According to the codes, PGA values of 0.4g, 0.3g, 0.2g and 0.1g are assigned for zones ranging from I to IV, respectively. However, in seismic hazard modeling studies for Turkey, some investigators (e.g. Erdik et al., 1985; Gu lkan et al., 1993; Kayabali, 2002) employed the relationships suggested for California, specifically by Joyner and Boore (1981, 1988) due to some similarities between the San Andreas Fault in USA and the North Anatolian Fault Zone in Turkey. The seismic zoning map of Turkey have the same setbacks and these are outlined by Kayabali and Akin (2002). In order to minimize the effects of the setbacks in the present map, new maps were constructed by Kayabali (2002) and Kayabali and Akin (2002) using the probabilistic and deterministic approaches, respectively. However, Kayabali and Akin (2002) also indicated that the probabilistic-based maps appear to have employed relatively large seismic zones, and therefore, deterministic-based seismic map seems to be more realistic. In the construction of the deterministic based isoacceleration and seismic zoning maps of Turkey, Kayabali and Akin (2002) considered only active faults and ignored locations of the epicenters. In addition, they compared two domestic attenuation equations of strong ground motion for the earthquakes of Turkey developed by Inan et al. (1996) and Aydan

et al. (1996), and some imported equations. Based on their comparisons, they decided to use the equation developed by Sadigh et al. (1997) as the appropriate prediction equation for the construction of their map. However, the equation by Sadigh et al. (1997) does not account for normal faulting, and it was a nonsense comparison through the use of the distance to epicenter instead of hypocentral distance when they considered the equation, which was developed by Aydan et al. (1996) modified by Aydan (2001) for Turkey. The PGA values picked up from this isoacceleration map are for bedrock and it is necessary to carry out site response analyses for the computation of maximum PGA for soil sites. Three domestic attenuation relationships of PGA for the earthquakes of Turkey were suggested by Inan et al. (1996), Aydan (2001), and Gu lkan and Kalkan (2002). The PGA equation developed by Inan et al. (1996) is as follows: log PGA 0:65M 0:9 log R 0:44 1

where M is the earthquake magnitude and R is the distance to epicenter in kilometers. No distinction is considered between the records obtained from the stations founded on rock and soil sites in this attenuation relationship. It is also noted that type of the magnitude employed in the equation and data selection criteria have not been mentioned in the related literature. On the other hand, this relationship yields unusually high values of PGA particularly in near source areas. The second attenuation relationship based on a large database system, which is called TURDIVAZ and involved recordings at stations on soil and rocky grounds operated by ERD, KOERI and ITU for the characteristics of acceleration waves of Turkish earthquakes, was developed by Aydan and published in a report (Aydan, 1997) and in an article (Aydan et al., 1996), respectively. This attenuation relationship is given in the following form: amax 2:8e0:9Ms e0:025R 1 2

where amax is the maximum ground acceleration, and Ms and R are the surface magnitude and the hypocentral distance of a given earthquake, respectively. At the beginning the number of data employed for the above equation was 60. This function consists of two expo-

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

267

nential functions, which fundamentally govern the overall form of attenuation. Two criteria were selected by its originator in developing this equation: (i) when the magnitude is zero, there should not be any ground motion, and (ii) it should also be capable of estimating strong ground motions of large earthquakes of Turkey with known accelerations. However, Aydan recognized a small problem with Eq. (2), that is, the ground acceleration will have a negative value when R goes to infinity although it will almost no effect on the estimated maximum ground acceleration. Therefore, form of the equation was slightly modified in order to satisfy the condition, that is, the maximum ground acceleration should be nil when R goes to infinity, and re-written in the following form (Aydan, 2001). amax 2:8e0:9Ms 1e0:025R 3

The coefficient 2.8 appearing in Eq. (3) is considered for soils and reduced to 0.56 for firm soils and rocky grounds as the site condition coefficient. This attenuation relationship generally simulated the records of big earthquakes of Turkey. The most recent attenuation relationship for Turkey was developed by Gu lkan and Kalkan (2002) by using the same general form of the equation proposed for shallow earthquakes in Western North America by Boore et al. (1997). The ground motion parameter estimation is as follows: ln Y b1 b2 M 6 b3 M 62 b5 ln r bv lnVs =VA ;
2 0:5 r R2 cl h ;

4 5

where Y is the ground motion parameter (PGA or PSA in g), M is (moment) magnitude; Rcl is the closest horizontal distance between the recording station and a point on the horizontal projection of the rupture zone on the earths surface in km; Vs is the shear wave velocity for the station in m/s; b1, b2, b3, b5, h, bv and VA are the parameters to be determined. Here h is a fictitious depth and VA is a fictitious shear-wave velocity that is determined by regression. Gu lkan and Kalkan (2002) utilized 47 horizontal components of only main shocks of 19 earthquakes with magni-

tudes Mw z 5 occurred in Turkey between 1976 and 1999, and omitted the PGA values less than 40 gal. Half of the data they employed was from the devastating Kocaeli and Du zce earthquakes of 1999. Based on the Kocaeli and Du zce events, Gu lkan and Kalkan (2002) compared their equations to some imported attenuation relationships not specifically from recordings in Turkey, and concluded that the imported relationships overestimate the peak and spectral acceleration values for up to about 15 20 km, for larger distances the reverse holds. These investigators also recommended that as additional strong motion records, shear-wave velocity profiles for recording sites and better determined distance data become available for Turkey, their attenuation relationship can be progressively modified and improved, and its uncertainties reduced. On the basis of the abovementioned information and brief evaluation, the authors of the present paper believe that the attenuation relationships derived in other countries and for different tectonic regimes should be carefully utilized for seismic assessments in Turkey. Therefore, the authors considered that derivation of an attenuation equation for PGA based on a larger database as a contributory study to those carried out to develop domestic equations for Turkish earthquakes, and its use in practice may become useful. In this study, an attempt was made to derive an attenuation equation of PGA for rock, soil and soft soil sites in Turkey. The database, employed in this study included the records from the earthquakes of Mw z 4 between 1976 and November 2003. Among all the three-component records, 221 records of 122 earthquakes were selected for regression analysis, and effects of the site conditions were also considered. From the regression analysis of data, the equation to predict PGA for the sites underlain by rock, soil and soft soil were established. The PGA values estimated from the equation developed in this study and those from some previous domestic attenuation equations and some imported models based on worldwide data were compared. In addition, using the proposed attenuation equation and considering the epicenters of the earthquakes and the known active faults of Turkey, PGA values were calculated. Then the calculated PGA values were contoured to obtain an iso-acceleration map of Turkey. The PGA value picked up from the map

268

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

for any site is for bedrock, and when it is multiplied by the coefficients, the PGA of soil and soft soils can be found.

2. Data selection criteria and database Installation of the acceleration recorders was initiated in Turkey in 1973 in relation with the project of Strong Ground Motion Network of Turkey, under the responsibility of the Earthquake Research Department (ERD) of the General Directorate of Disaster Affairs (GDDA). Since that date, the strong motion network has grown considerably, and the first strong motion recording of an earthquake was obtained in Denizli on 19 August 1976, western Turkey. While the total number of instruments was 120 up to 2001, at the end of 2002 the total number of the instruments reached to 163 (ERD, 2003). Ninety-six of these accelerometers are digital, while 67 instruments are analog. The accelerometric sites in Turkey are generally located along the North Anatolian and East Anatolian Fault Zones and as well as in the southwestern Anatolia. In addition to these instruments forming the network, a limited number of temporary stations installed by the Kandilli Observatory and Earthquake Research Institute of Bogazic i University (KOERI), stanbul Technical University (ITU), USGS, LomontY

Doberty Earth Observatory (LDEO, USA) and Universite Joseph Fourier after the devastating 1999 Kocaeli Earthquake are also present. The records of these stations were downloaded from the web sites of the Consortium of Organizations for Strong Motion Observation Systems (COSMOS, 2003) and USGS. Some of the accelerometric data, available on the Internet, do not include corresponding earthquake characteristics (location and magnitude), and there are some contradictions between different databases. In order to decrease the number of unknowns and uncertainties, and consequently improve the reliability of the derived attenuation equation, additional data associated with the earthquakes occurred in Turkey were also found from the Internet sites of ETHZ (2003), ISESD (2003), USGS-NEIC (2003), ISC (2003) and HARVARD (2003). Among all-the threecomponent accessible records between 1976 and November 2003, 221 records from 122 earthquakes of Turkey were selected based on the criteria outlined in the following paragraphs. Locations of the accelerometric sites in Turkey corresponding to the data selected in this study are shown in Fig. 1. Exclusion of records based on minimum PGA has been proposed as selection criteria and they are reviewed by Douglas (2003). In this study, the acceleograms with a PGA z 20 gal were selected considering the criterion by Campbell (1981) to avoid bias in

Fig. 1. The locations of the strong ground motion stations of Turkey employed in this study.

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

269

trigger threshold. The PGA data used in the analysis were mostly from the stations founded in small buildings, and the larger horizontal component of each record was selected. The best scale for scientific and engineering purposes is the moment magnitude (Mw) scale since it is related to the rupture parameters. Therefore, in this study earthquake size was characterized by Mw. Because smaller earthquakes are generally not of engineering significance, the earthquakes with Mw z 4 were considered. However, the magnitudes of the earthquakes occurred in Turkey are reported by dif-

ferent institutions in various scales. Based on the database for a total of 170 events in Turkey occurred between 1976 and November 2003, the numbers of the magnitudes given by different institutions in Mw, Ms, Mb, Md and ML scales are 96, 95, 150, 75 and 69, respectively. In other words, Mw values are not available for all events. Therefore, it was decided to derive moment magnitude for all records, to provide a uniform and reliable scale for the attenuation relationship, which was developed in this study. For the purpose, values of Mw (from ETHZ and Harvard) were correlated to Ms (from ETHZ, ISC, USGS,

Fig. 2. Correlations between the reported Mw, and the reported Ms, Mb, Md and ML values for Turkish earthquakes (r: correlation coefficient; S.D.: standard deviation).

270

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Harvard and ISESD), Mb (from ETHZ, ISC, USGS and Harvard), Md and ML (from ERD) values. Using data sets of 54, 92, 28 and 27 between Ms Mw, Mb Mw, Md Mw and ML Mw, respectively, the relationship and conversion equations derived between Mw and other magnitude scales are given in Fig. 2. These relationships yielded considerably high correlation coefficients greater than 0.9. Of the 122 earthquakes considered in this study based on the selection criteria, magnitudes of 49 earthquakes have been directly presented in Mw scale. Therefore, magnitudes in Mw scale for the rest (73 earthquakes) were derived from the equations given in Fig. 2 (39, 18, 10 and 6 Mw values from Ms, Mb, Md and ML, respectively). The distance to epicenter is the easiest measure to use because the epicenter is the location information given for all earthquakes. For small earthquakes, the use of distance to epicenter in hazard analysis is reasonably straightforward because easily available catalogues of previous epicenters can be used as the future sources or if line or surface source zones are used then epicenters can be distributed on these source zones (Douglas, 2003). For large-magnitude earthquakes, the closest distance measures are gen-

erally preferred over the point source distances, such as distance to the surface projection of the rupture (e.g. Joyner and Boore, 1981) or rupture distance (Campbell, 1981), at least for records from earthquakes with Mw>6.5. However, for most of the events, particularly for small events that have occurred in Turkey, rupture surfaces have not been defined clearly, and these distances are more difficult to estimate. One of the other distance measures that is available for most earthquakes is hypocentral distance (Rh). However, accurate measures of focal depth are often difficult, and therefore, estimation of hypocentral distance is affected from this limitation. Most damaging earthquakes occur within a shallow region of the crust (about the top 30 km) and hence Rh and distance to epicenter (Rc) become equal at intermediate and large distances (Douglas, 2003). It is also noted that the values of focal depth reported by various institutions can be different as shown in Fig. 3 for some earthquakes with Mw>5.5 from Turkey. In this study, due to the abovementioned reasons, distance to epicenter, Re, is preferred to be used as site source distance in PGA estimation relation. Minimum and maximum distance criteria

Fig. 3. Comparison of the depth values of some selected earthquakes of Turkey (Mw>5.5) reported by various institutions (earthquake numbers refer to the earthquakes given in Table 1).

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

271

are sometimes applied. A minimum distance criterion of 2 km was applied by Wang et al. (1999) because 2 km is the minimum error in epicentral locations and hence including records from smaller distances may give errors in the results. As mentioned by Douglas (2003), only records associated with reliable measures were used by some investigators (Campbell, 1981; Sabetta and Pugliese, 1987) by including only earthquakes with locations (epicenters or rupture distance) known to within 5 km or less. On the other hand, in the majority of the strong ground motion relations suggested for tectonically active regions (e.g. Boore et al., 1997; Campbell, 1997; Sadigh et al., 1997) the upper bound for site source distance is taken 100 km, which is the range where ground motions have engineering significance. Therefore, 5 and 100 km were taken as the lower and upper bounds of the distance to epicenter, respectively, and the records, for which the distance to epicenter does not fall into this range were omitted. One of the extremely difficult items in determining the site condition coefficient of the attenuation relationships is the ground conditions. Local site conditions at an acceleograph station can affect the strong motion recorded. Therefore, attempts have been made in most ground estimation relations to model the effect of nearsurface ground conditions or strong motion. Data selection criteria, which seek to limit the acceleograms used to those recorded at stations with similar local site conditions, are the simplest techniques. While the widely accepted method quantitatively define the near-surface material based on shear-wave velocity, Vs, beneath the station. However, without consistent site classifications for the attenuation relations, it is often difficult to know how to apply the relations to a specific site. But information on Vs is currently lacking for the stations in Turkey. Based on their experience and Bard (2002) classified with the Iranian data, Zare the records from Turkey for site conditions according to the frequency band of the fundamental frequency based on their H/V ratio. This ratio was chosen based on the location of the broadened H/V spectral ratios and the value of its amplitude. These investigators suggest that this is the only method that may reveal the site response, in the absence of any reliable geological and geotechnical data, such as Vs values. In and Bard (2002) divided soil groups for brief, Zare

Turkey in ascending order for Vs and fundamental frequency ( fo) site category: 1rock and hard alluvial soil corresponds to Vs>800 m/s, fo>15 Hz, site category 2alluvial sites (500 < Vs < 700 m/s, 5 < fo < 15 Hz), site category 3(300 < Vs < 500 m/s, 2 < fo < 5 Hz), and site category 4 relates to soft soil (Vs < 300 m/s, fo < 2 Hz). On the other hand, Gu lkan and Kalkan (2002), who also suggested an attenuation modeling of recent earthquakes in Turkey, indicated that because the actual shear-wave velocity and detailed site description are not available for most stations in Turkey, they obtained the type of geologic material underlying each recording site in a number of ways, such as consultation with geologist at ERD, various geologic maps, past earthquake reports and geological references prepared for Turkey. In the light of this information, they divided soil groups for Turkey into three, such as rock, soil and soft soil. Site conditions at the strong ground motion stations operated by the ERD (Inan et al., 1996) are grouped as soil (old deposits), alluvium (soft deposits such as recent, coastal and beach deposits) and rock based on the observations on the sites during the installation of the devices and geological references. However, this categorization did not consider measured Vs values. Similarly, site conditions at some more recent stations, generally established and operated by KOERI after the devastating 1999 Kocaeli earthquake of Turkey, were defined by visual inspections and grouped into three categories (rock, soil and soft soil). In addition, detailed site conditions at five stations established after the 1995 Dinar earthquake for aftershock studies at Dinar are available in the published literature (Durukal et al., 1998). The other useful sources for the site conditions are available from the Internet sites of ISESD and USGS. Site conditions in the database reported from these institutions are updated at different times based on new information and constituted of three groups as rock, soil, soft soil. The database reported by ISESD indicates that there are only seven stations in Turkey, where site conditions were categorized on shear-wave velocity measurements. Consequently, it was considered that the general trend among different institutions is to categorize site conditions in Turkey in three groups. In this study, therefore, three site conditions, namely rock, soil and soft soil sites, were considered. For

272

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Table 1 Strong ground motion records used in the development of the attenuation relationship for peak ground acceleration (PGA) for Turkey No. Station Code Station coordinates Lat. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Denizli Meteoroloji Mu d. Cerkes Meteoroloji Mu d. Izmir Meteoroloji Mu d. Malatya Meteoroloji Mu d. Muradiye Meteoroloji Mu d. Dursunbey Kandilli Go zlem Ist. Hatay Bayindirlik ve Iskan Mu d. Go nen Meteoroloji Mu d. Edincik Kandilli Go zlem Ist. Tekirdag Meteoroloji Mu d. Edremit Meteoroloji Mu d. Balikesir Meteoroloji Mu d. Horasan Meteoroloji Mu d. Erzurum Meteoroloji Mu d. Foca Gu mru k Mu d. Kigi Meteoroloji Mu d. Ko ycegiz Meteoroloji Mu d. Go lbasi Devlet Hastanesi Kusadasi Meteoroloji Mu d. Go lbasi Devlet Hastanesi Malatya Meteoroloji Mu d. Muradiye Meteoroloji Mu d. Foca Gu mru k Mu d. Istanbul Bayindirlik ve Isk. Mu d. Amasya Meteoroloji Mu d. Erzincan Meteoroloji Mu d. Tercan Meteoroloji Mu d. Refahiye Kaymakamlik Binasi Erzincan Meteoroloji Mu d. Erzincan Eksisu Izmir Meteoroloji Mu d. Kusadasi Meteoroloji Mu d. Ilica Meteoroloji Mu d. Cesme Meteorological Station Islahiye Meteoroloji Mu d. Foca Gu mru k Mu d. Foca Gu mru k Mu d. Ilica Meteoroloji Mu d. Ko ycegiz Meteoroloji Mu d. Ko ycegiz Meteoroloji Mu d. Ko ycegiz Meteoroloji Mu d. Ko ycegiz Meteoroloji Mu d. Tercan Meteoroloji Mu d. Van Bayindirlik ve Iskan Mu d. Tekirdag Meteoroloji Mu d. DNZ CER IZM MLT MUR DUR HTY GNN EDC TKR EDR BLK HRS ERZ FOC KOY GOL KUS GOL MLT MUR FOC IST AMS ERC TER REF ERC SERE IZM KUS ILI CES ISL FOC FOC ILI KOY KOY KOY KOY TER VAN TKR 37.814 40.88 38.4 38.35 39.03 39.67 36.25 40.08 40.36 40.96 39.61 39.66 40.04 39.906 38.64 39.34 36.97 37.781 37.861 37.781 38.35 39.03 38.64 41.08 40.63 39.752 39.777 39.901 39.752 39.733 38.4 37.861 38.31 38.333 37.05 38.64 38.64 38.31 36.97 36.97 36.97 36.97 39.777 38.504 40.96 Lon. 29.112 32.91 27.19 38.34 43.7 28.53 36.11 27.68 27.89 27.53 27.03 27.86 42.17 41.256 26.77 40.28 28.694 37.641 27.266 37.641 38.34 43.7 26.77 29.09 35.87 39.487 40.391 38.769 39.487 39.783 27.19 27.266 26.31 26.317 36.6 26.77 26.77 26.31 28.694 28.694 28.694 28.694 40.391 43.406 27.53 19.08.1976 05.10.1977 09.12.1977 21.09.1978 11.04.1979 18.07.1979 Earthquake Epicenter (date) coordinates Lat. 37.71 41.01 38.35 37.97 39.12 39.66 Lon. 29 33.57 27.23 38.59 43.91 28.65 35.89 27.21 27.21 27.21 27.21 27.21 42.18 42.18 25.68 39.77 28.85 37.79 27.39 37.91 37.91 44.12 27 29.09 35.8 39.63 39.63 39.63 39.93 39.93 26.99 26.99 26.99 27 35.84 26.54 26.6 26.6 28.89 28.82 28.8 28.87 40.99 43.33 27.67 Re PGA (gal) (km) NS 15.2 348.53 57.2 36.03 6.6 158.91 47.5 14.08 20.7 46.04 10.3 232.29 21.6 154.05 48.6 57.7 75.0 81.4 92.7 34.4 92.7 97.8 80.6 13.8 29.6 15.5 34.7 53.2 37.3 31.5 31.1 50.11 53.44 29.89 25.38 22.55 173.3 35.49 24.17 163.06 103.24 114.7 55.52 68.54 23.57 49.5 31.61 27.58 EW UD 2 3 2 1 1 2 2 2 1 1 2 2 2 1 1 2 2 1 2 1 1 1 1 2 3 2 2 2 2 3 2 2 2 2 2 1 1 2 2 2 2 2 2 2 1 5.5a 5.8 5.2a 5.0a 5.0a 5.3 4.6a 6.1 6.1 6.1 6.1 6.1 6.6 6.6 5.5b 4.9a 5.1a 6.0 4.1c 5.8 5.8 5.5 4.7c 5.3a 5.1b 6.6 6.6 6.6 5.9 5.9 6.0 6.0 6.0 6.0 5.3a 5.5 5.5a 5.5a 5.4a 4.7c 5.4 4.3d 5.2 5.2b 5.0a Site Mw C.

290.36 173.29 38.94 16.18 272.97 87.25 35.79 21.12 45.22 24.67 288.25 199.77 135.6 46.77 46.51 34.91 27.78 20.71 150.26 24.99 23 89.1 114.46 76.04 94.43 34.42 24.81 51.18 41.4 18.21 144.32 37.68 31.67 17.19 17.47 24.72 87.92 31.94 23.52 42.63 68.59 38.96 54.16 18.01 26.04 20.65 28.32 9.68

30.06.1981 36.17 05.07.1983 05.07.1983 05.07.1983 05.07.1983 05.07.1983 30.10.1983 30.10.1983 17.06.1984 12.08.1985 06.12.1985 05.05.1986 01.06.1986 06.06.1986 06.06.1986 20.04.1988 04.08.1988 12.02.1991 12.02.1992 13.03.1992 13.03.1992 13.03.1992 15.03.1992 15.03.1992 06.11.1992 06.11.1992 06.11.1992 06.11.1992 03.01.1994 24.05.1994 24.05.1994 24.05.1994 13.11.1994 13.11.1994 13.11.1994 13.11.1994 29.01.1995 26.02.1995 40.33 40.33 40.33 40.33 40.33 40.35 40.35 38.87 39.95 36.97 38.02 37.96 38.01 38.01 39.11 38.86 40.8 40.58 39.72 39.72 39.72 39.53 39.53 38.16 38.16 38.16 38.16 37 38.66 38.76 38.76 36.97 37 36.96 37.17 39.98 38.6

8.1 37.1 12.7 470.92 65.3 39.38 76.2 67.21 45.2 25.8 31.8 41.1 61.6 62.6 67.6 20.1 19.9 56.0 17.4 11.7 9.5 27.1 55.8 12.5 16.9 32.45 30.49 83.49 16.65

29.03 24.94 404.97 238.55 26.97 22.67 85.93 31.57 39.3 112.5 38.34 71.8 37.81 28.3 19.43 49.8 46.84 26.88 96.51 20.78 39.95 23.78 48.52 15.5 45 18.47 40.7 21.19 62.1 25.77 14.9 19.14 29.6 25.6 14.6 57.91 20.63 34.25 20.08 24.81 15.5 11

20.57 36.06 57.65 24.91 72.79 26.6 57.16 25.54 44.98 28 38

13.04.1995 40.85

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 Table 1 (continued) No. Station Code Station coordinates Lat. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 Tekirdag Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Cardak Saglik Ocagi Burdur Meteoroloji Mu d. Denizli Meteoroloji Mu d. Dinar Meteoroloji Mu d. Cardak Saglik Ocagi Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Meteoroloji Mu d. Dinar Cezaevi Dinar Jandarma Karakolu Dinar Devlet Hastanesi Dinar Ko y Hizmetleri Erzincan Bayindirlik ve Isk Mu d. Erzincan Bayindirlik ve Isk Mu d. Erzincan Bayindirlik ve Isk Mu d. Kusadasi Meteoroloji Mu d. Kusadasi Meteoroloji Mu d. Tekirdag Meteoroloji Mu d. Osmancik Belediye Binasi Amasya Bayindirlik ve Isk Mu d. Amasya Bayindirlik ve Isk Mu d. Merzifon Meteoroloji Mu d. Buldan Kaymakamlik Binasi Hatay Bayindirlik ve Iskan Mu d. Hatay Bayindirlik ve Iskan Mu d. Amasya Bayindirlik ve Isk. Mu d. Sakarya Bayindirlik ve Isk. Mu d. Gelibolu Karayollari Mu d. Edremit Meteoroloji Mu d. Dinar Meteoroloji Mu d. Cardak Saglik Ocagi Horasan Meteoroloji Mu d. Elazig Bayindirlik Mu d. Ceyhan Tarim Ilce Mu d. TKR DIN DIN DIN DIN DIN DIN CRD BRD DNZ DIN CRD DIN DIN DIN DCE DJK DDH DKH ERC ERC ERC KUS KUS TKR OSM AMS 40.96 38.06 38.06 38.06 38.06 38.06 38.06 37.825 37.723 37.814 38.06 37.825 38.06 38.06 38.06 38.075 38.069 38.067 38.053 39.743 39.743 39.743 37.861 37.861 40.96 40.97 40.63 Lon. 27.53 30.155 30.155 30.155 30.155 30.155 30.155 29.668 30.294 29.112 30.155 29.668 30.155 30.155 30.155 30.161 30.16 30.171 30.139 39.512 39.512 39.512 27.266 27.266 27.53 34.83 35.87 35.87 35.49 28.834 36.11 36.11 35.87 30.384 26.67 27.016 30.155 29.668 42.173 39.193 35.81 18.04.1995 26.09.1995 26.09.1995 27.09.1995 28.09.1995 01.10.1995 01.10.1995 01.10.1995 01.10.1995 01.10.1995 01.10.1995 01.10.1995 03.10.1995 05.10.1995 06.10.1995 11.10.1995 11.10.1995 11.10.1995 11.10.1995 05.12.1995 Earthquake Epicenter (date) coordinates Lat. 40.8 38.04 38.09 38.11 38.56 38.04 38.11 38.11 38.11 38.11 38.1 38.1 38.01 38.04 38.03 38.12 38.12 38.12 38.12 39.3 Lon. 27.84 30.03 30.01 30.02 30.47 30.09 30.05 30.05 30.05 30.05 30.02 30.02 30.05 30.1 30.09 30.18 30.18 30.18 30.18 40.3 40.1 39.23 27.13 26.64 27.45 35.29 35.29 35.23 35.23 28.92 36.12 36.33 35.3 30.42 26.43 26.7 30.04 30.04 41.61 38.94 35.55 31.5 11.2 13.1 13.0 61.9 6.1 10.7 46.0 48.0 88.5 12.6 43.4 10.7 5.3 6.6 5.3 5.9 5.9 8.3 83.5 70.4 28.3 44.8 55.7 19.0 46.3 50.3 56.7 PGA (gal) Re (km) NS 25.5 106 54.26 86.76 39.62 91.33 281.63 65.07 41.61 20 224.66 24.83 68.53 104.32 98.85 44.74 31.31 40.53 86.24 28.27 27.78 47.91 15.45 21.33 10.5 15.65 26.5 20 EW UD 1 3 3 3 3 3 3 2 2 2 3 2 3 3 3 2 2 1 2 2 2 2 2 2 1 2 3 3 2 2 2 2 3 1 2 2 3 2 2 2 3

273

Site Mw C. 4.9c 5.0a 5.1a 5.0a 4.2c 4.3c 6.4 6.4 6.4 6.4 5.4a 5.4a 5.0a 5.2b 5.0b 4.3a 4.3a 4.3a 4.3a 5.8 5.9a 4.2d 4.7b 5.4 4.1c 5.7 5.7 5.6 5.6 5.2 5.7 4.1c 5.4a 4.3a 5.2a 5.0a 5.2 5.2 5.2 5.1 6.3

34.5 8 182.73 75.56 81.02 48.77 180.38 71.74 42.72 12.11 171.79 37.66 329.72 150.68 61.3 98.47 43.92 33.86 10 10 125.87 54.85 20.87 13.18 145.59 97.68 128.84 80.48 168.08 44.57 41.14 15.93 63.6 13.18 25.15 13.31 47.18 14.53 24.02 23.99 22.85 38.47 21.35 33.44 21 30.88 54 33.5 102.34 38.51 149 19 21 71.59 19.43 20.06 130.9 27.88 32.5 15 223.28 17.64 35.15 13.36 22.46 7 11.62 25.5 16.5 28.53 28.02 89.5 16 14.5 15.14 8.75 14.17 27.71 19.15 23.5 17.5 86.47

05.12.1995 39.3 14.02.1996 39.61 20.02.1996 02.04.1996 14.04.1996 14.08.1996 14.08.1996 38.25 37.78 40.8 40.74 40.74

AMS 40.63 MRZ 40.88 BLD 38.045 HTY 36.25 HTY 36.25

14.08.1996 40.79 14.08.1996 40.79 21.01.1997 38.12 22.01.1997 36.14 23.01.1997 36.16 28.02.1997 40.68 21.10.1997 40.7 25.10.1997 05.03.1998 04.04.1998 04.04.1998 13.04.1998 09.05.1998 27.06.1998 40.49 39.68 38.14 38.14 39.91 38.38 36.85

24.0 33.38 11.2 24.37 12.2 134.5 22.1 48.3 5.3 27.5 21 33.87

AMS 40.63 SKR GLB EDR DIN CRD HRS ELZ CYH 40.739 40.43 39.583 38.06 37.825 40.043 38.672 37.05

21.3 42.42 29.1 27.11 13.4 134.73 47.8 24.48 50.1 38.5 39.2 25.5 32.0 273.55

(continued on next page)

274 Table 1 (continued) No. Station

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Code Station coordinates Lat. Lon. 35.367 34.65 36.16 36.6 35.617 35.455 35.633 34.65 35.633 35.455 27.229 42.173 39.193 34.037 41.262 29.762 30.384 29.691 29.44 29.3607 29.087 30.734 29.09 29.0111 28.95 29.068 29.131 28.8199 29.019 28.997 28.908 28.928 28.849 29.907 30.622 31.003 29.762 30.384 30.257 30.38 29.79347 29.78836 29.78965

Earthquake Epicenter (date) coordinates Lat. 27.06.1998 36.85 27.06.1998 36.85 27.06.1998 36.85 27.06.1998 36.85 04.07.1998 36.84 04.07.1998 36.84 04.07.1998 04.07.1998 04.07.1998 04.07.1998 09.07.1998 19.12.1998 13.04.1999 16.06.1999 20.07.1999 36.84 36.84 36.88 36.88 38.08 39.82 38.53 40.96 39.53 Lon. 35.55 35.55 35.55 35.55 35.44 35.44 35.44 35.44 35.62 35.62 26.68 42.13 39.25 33.87 41.28 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 29.97 30.7 30.7 29.92 29.92 29.92 29.92 29.92 29.92 29.92

PGA (gal) Re (km) NS 36.0 28.5 80.0 119.29 89.3 27.07 95.8 21.35 16.1 24.1 26.8 122.1 27.4 70.2 17.0 26.7 63.4 25.0 16.5 15.2 41.4 28.3 46.7 22 90 27 26.5 12.5 34.36 33.51 EW 33.1 132.12 25.79 18.22 20 132.9 20.8 60.9 23.1 90.7 12.5 19.5 31 30.49 29.88 UD 19.74 22.05 12.36 14.11 19.6 35.9 13.8 12.1 12.6 25 5.6 14 9.5 17.76 23.5

Site Mw C.

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129

Karatas Meteoroloji Mu d. Mersin Meteoroloji Mu d. Hatay Bayindirlik ve Iskan Mu d. Islahiye Meteoroloji Mu d. Nacarli Ko yu Kilicli Ko yu gretim Ilko Okulu Baklal Ko yu Saglik Ocagi Mersin Meteoroloji Mu d. Baklali Ko yu Saglik Ocagi Kilicli Ko yu gretim Ilko Okulu Bornava Ziraat Faku ltesi Horasan Tarim Ilce Mu d. Elazig Bayindirlik Mu d. Tosya Meteoroloji Mu d. Erzurum Bayindirlik ve Isk. Mu d. Yarimca Petkim Tesisleri Sakarya Bayindirlik ve Isk. Mu d. Iznik Karayollari Sefligi Gebze Tu bitak Marmara A. M. Darica Arcelik Arge Binasi Heybeliada Sanatoryum Go ynu k Devlet Hastanesi Istanbul Bayindirlik ve Isk. Mu d. Levent Yapi Kredi Plaza Fatih Tu rbesi Bursa Tofas Fabrikasi Bursa Sivil Savunma Mu d. Yesilko y Havalimani ITU Istanbul Maslak ITU Istanbul Mecidiyekoy ITU Istanbul Zeytinburnu ITU Istanbul K.M.Pasa ITU Istanbul Atakoy Izmit Meteoroloji Istasyonu Akyazi Orman Isletme Mud. Golyaka Jandarma Kislasi Yarimca Petkim Sakarya Bayindirlik ve Isk. Mu d. Sapanca Saglik Ocagi USGS Golden Station tyw USGS Golden Station KOR USGS Golden Station TUN USGS Golden Station TUS

KRT 36.561 MRS 36.83 HTY 36.213 ISL 37.05 NAC 36.87 KIL 37.081 BKL MRS BKL KIL BRN HRS ELZ TOS ERZ SKR IZN GBZ HAS GYN IST FAT BRS MSK MCD ZYT KMP ATK IZT . GLY SKR SPN 37.033 36.83 37.033 37.081 38.455 40.043 38.672 41.013 39.903 40.7639 40.737 40.437 40.82 40.8236 40.8688 40.385 41.08 41.0811 41.0196 40.2605 40.184 40.9823 41.104 41.065 40.986 41.003 40.989 40.761 40.67 40.799 40.7639 40.737 40.688 40.7 40.76519 40.75704 40.75455

2 3 2 2 1 3 3 3 3 3 3 2 2 2 1 3 1 3 2 2 1 2 2 1 3 3 2 2 1 2 3 3 3 1 3 3 3 1 3 3 1 2 2

6.3 6.3 6.3 6.3 5.4 5.4 5.4 5.4 4.6a 4.6a 5.0a 4.9a 4.5a 4.1a 4.8a 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 5.3a 5.3a 5.1 5.1 5.1 5.1 5.1 5.1 5.1

17.08.1999 40.76 17.08.1999 40.76 17.08.1999 40.76 17.08.1999 40.76 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 17.08.1999 22.08.1999 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.76 40.69

17.5 230.22 34.9 42.9 91.89 45.1 284.82 51.7 211.37 75.2 56.15 76.7 117.9 81.9 60.67 88.0 41.08 90.4 189.39 94.2 100.89 95.5 54.32 99.7 90.21 88.5 88.4 92.6 91.6 97.5 5.3 7.0 33.6 28.3 13.4 39.1 29.3 39.5 10.9 11.2 11.1 27.5 20.8 24.41 30.5 23.50 36.34 46.34 33.76

322.2 407.04 259 123.32 141.45

133.68 110.23 137.7 129.9 42.66 36.22 35.52 161.87 100.04 45.81 25.73 84.47 47.7 27.1 61.8 27.5 112 46.8 128.3 84.2 161.1 61.6 219.2 139.2 35 19.9 21.4 14 17.06 31.7 24.69 32.08 56.97 60.33 9.9 10.1 8.42 24.4 14.66 14.39 24.18 22.89

22.08.1999 40.69 31.08.1999 40.75 31.08.1999 40.75 31.08.1999 31.08.1999 31.08.1999 31.08.1999 31.08.1999 40.75 40.75 40.75 40.75 40.75

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 Table 1 (continued) No. Station Code Station coordinates Lat. 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 USGS Golden Station TUW USGS Golden Station TYN USGS Golden Station TUN USGS Golden Station TUS USGS Golden Station TUW USGS Golden Station KOR USGS Golden Station TUW Sakarya Bayindirlik ve Isk. Mu d. Yarimca Petkim Heybeliada Senatoryum Tepetarla Koyu Muhtar Evi Bahcecik Seymen Kislasi USGS Golden Station KOR USGS Golden Station SUB USGS Golden Station DOR USGS Golden Station OIL USGS Golden Station TUW USGS Golden Station TYN USGS Golden Station TYR USGS Golden Station TYW USGS Golden Station TYR Sakarya Bayindirlik ve Isk. Mu d. USGS Golden Station TYR Bahcecik Seymen Kislasi Sapanca Saglik Ocagi USGS Golden Station DOR USGS Golden Station KOR USGS Golden Station OIL USGS Golden Station TUW USGS Golden Station TYW USGS Golden Station SUB USGS Golden Station YAR LDEO Station No. C0362 CH LDEO Station No. C1061 LDEO Station No. C1060 BU Sakarya Bayindirlik ve Isk. Mu d. Yarimca Petkim LDEO Station No. C1060 BU LDEO Station No. C1061 Sakarya Bayindirlik ve Isk. Mu d. Du zce Meteoroloji Mu d. 40.75646 40.71192 40.75704 40.75455 40.75646 40.76517 40.75644 40.737 40.7639 40.8688 40.72 40.71 40.76505 40.6869 40.77736 40.75551 40.75644 40.71190 40.73701 40.70923 40.73702 40.737 40.73705 40.71 40.688 40.77736 40.76512 40.75549 40.75644 40.70924 40.6869 40.64472 40.67 40.72 40.777 40.737 40.7639 40.777 40.72 40.737 40.85 Lon. 29.78672 30.39047 29.78836 29.78965 29.78671 29.79343 29.78671 30.384 29.762 29.087 30.079 29.907 29.79332 29.49396 29.51404 29.78091 29.78668 30.39040 30.38008 30.39163 30.38009 30.384 30.38010 29.907 30.257 29.51404 29.79343 29.78091 29.78673 30.39165 29.49396 29.27490 30.666 30.792 30.613 30.384 29.762 30.613 30.792 30.384 31.17 31.08.1999 31.08.1999 31.08.1999 31.08.1999 31.08.1999 06.09.1999 06.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 13.09.1999 17.09.1999 19.09.1999 24.09.1999 29.09.1999 29.09.1999 29.09.1999 29.09.1999 29.09.1999 29.09.1999 29.09.1999 29.09.1999 20.10.1999 07.11.1999 Earthquake Epicenter (date) coordinates Lat. 40.75 40.75 40.73 40.73 40.73 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.77 40.64 40.77 40.74 40.74 40.74 40.74 40.74 40.74 40.74 40.74 40.83 40.7 Lon. 29.92 29.92 29.95 29.95 29.95 29.73 29.73 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.1 30.13 30.52 30.23 29.33 29.33 29.33 29.33 29.33 29.33 29.33 29.33 29.03 30.72 30.72 30.72 30.2 30.2 30.2 30.2 30.3 31.21 11.4 40.3 14.1 13.9 14.2 5.4 5.1 24.2 28.5 85.9 5.9 17.6 26.1 52.5 49.9 27.2 26.7 25.6 24.2 25.8 21.6 15.7 PGA (gal) Re (km) NS 46.64 17.71 23.46 34.40 24.31 57.18 46.16 42.21 86 19.2 595.8 300.4 91.10 29.40 20.61 151.51 107.24 56.24 47.29 46.41 46.97 18.68 EW 36.89 19.89 32.44 56.47 30.20 24.62 44.37 50.6 UD 26.47 9.40 14.71 15.66 19.14 22.71 38.72 23.35 2 3 2 2 2 1 2 1 3 1 3 2 1 1 1 2 2 3 1 3 1 1 1 2 3 1 1 2 2 3 1 1 2 2 1 1 3 1 2 1 3

275

Site Mw C.

5.1 5.1 4.7a 4.7a 4.7a 4.1c 4.1c 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 4.2a 4.3a 4.3d 5.2a 5.2a 5.2a 5.2a 5.2a 5.2a 5.2a 5.2a 5.2a 5.0a 5.0a 5.0a 5.6 5.6 5.6 5.6 4.1c 7.2

. . . . . . SKR BHC . . . . . SKR . . . SKR DZC

89.9 50 35.4 31.6 332.9 186.8 379.4 70.7 108.40 58.30 42.00 18.60 23.29 25.47 164.09 54.42 110.07 55.71 52.41 21.63 50.98 29.32 47.20 24.86 70.51 19.76 32.53 11.41 26.03 91.5 17.6 182.35 109.65 208.36 125.71 26.65 96.40 30.94 48 138.5 25.2 15.96 21.6 10.2 70.88 21.08 28.68 26.67 8.77 76.30 12.06 63.1 64.3 24.4

13.3 15.46 49.0 74.2 78.8 21.2 16.3 200.24 39.6 58.59 38.5 219.49 39.0 92.98 90.7 18.17 15.2 109.70 29.6 23.49 5.7 51.9 6.5 12.5 62.7 35.3

07.11.1999 40.7 07.11.1999 40.7 11.11.1999 40.81 11.11.1999 40.81 11.11.1999 40.81 11.11.1999 40.81 11.11.1999 40.88 12.11.1999 40.74

17.5 206.54 37.2 35.2 51.2 17.4 17.1 31.4 44.4 39.795

345.28 133.33 23.9 41.6 50.8 78.49 19.5 44.4 20.1 20.08

12.7 407.69

513.78 339.64

(continued on next page)

276 Table 1 (continued) No. Station

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Code Station coordinates Lat. Lon. 31.182 31.61 30.384 30.876 30.792 30.666 30.613 31.61 31.61 31.61 31.61 31.61 31.61 30.384 31.015 31.61 31.015 30.792 30.384 31.149 31.61 31.149 31.149 30.384 29.114 32.833 30.622 30.384 31.149

Earthquake Epicenter (date) coordinates Lat. 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.74 12.11.1999 40.75 12.11.1999 40.75 12.11.1999 40.75 12.11.1999 40.74 12.11.1999 40.75 12.11.1999 40.74 13.11.1999 40.78 13.11.1999 40.78 13.11.1999 40.75 13.11.1999 40.83 13.11.1999 40.83 15.11.1999 40.91 20.12.1999 40.87 05.01.2000 40.84 20.01.2000 40.76 14.02.2000 40.9 02.04.2000 40.79 21.04.2000 37.85 06.06.2000 40.72 23.08.2000 40.68 23.08.2000 40.68 23.08.2000 40.68 Lon. 31.21 31.21 31.21 31.21 31.21 31.21 31.21 31.1 31.45 31.4 31.4 31.36 31.05 30.3 30.3 31.4 31.02 31.02 30.33 31.01 31.3 31.33 31.75 30.23 29.27 32.87 30.71 30.71 30.71

PGA (gal) Re (km) NS 30.8 120.99 33.7 739.51 69.5 17.33 EW 58.34 UD

Site Mw C.

171 Mudurnu Kaymakamlik Binasi 172 Bolu Bayindirlik ve Iskan Mu d. 173 Sakarya Bayindirlik ve Isk. Mu d. 174 LDEO Station No. C1058 BV 175 LDEO Station No. C1061 176 LDEO Station No. C0362 CH 177 LDEO Station No. C1060 BU 178 Bolu Bayindirlik ve Iskan Mu d. 179 Bolu Bayindirlik ve Iskan Mu d. 180 Bolu Bayindirlik ve Iskan Mu d. 181 Bolu Bayindirlik ve Iskan Mu d. 182 Bolu Bayindirlik ve Iskan Mu d. 183 Bolu Bayindirlik ve Iskan Mu d. 184 Sakarya Bayindirlik ve Isk. Mu d. 185 LDEO Station No. C1058 BV 186 Bolu Bayindirlik ve Iskan Mu d. 187 LDEO Station No. C1058 BV 188 LDEO Station No. C1061 189 Sakarya Bayindirlik ve Isk. Mu d. 190 Du zce Meteoroloji Mu d. 191 Bolu Bayindirlik ve Iskan Mu d. 192 Du zce Meteoroloji Mu d. 193 Du zce Meteoroloji Mu d. 194 Sakarya Bayindirlik ve Isk. Mu d. 195 Denizli Meteoroloji Mu d. 196 Cerkes Meteoroloji Mu d. 197 Akyazi Orman Isletme Mu d. 198 Sakarya Bayindirlik ve Isk. Mu d. 199 Du zce Meteoroloji Mu d.

BOL SKR . . . . BOL BOL BOL BOL BOL BOL SKR . BOL . . SKR DZC BOL DZC DZC SKR DNZ CER SKR DZC

40.463 40.747 40.737 40.743 40.72 40.67 40.777 40.747 40.747 40.747 40.747 40.747 40.747 40.737 40.755 40.747 40.755 40.72 40.737 40.844 40.747 40.844 40.844 40.737 37.812 40.814 40.67 40.737 40.844

1 3 1 2 2 2 1 3 3 3 3 3 3 1 2 3 2 2 1 3 3 3 3 1 2 3 3 1 3

7.2 7.2 7.2 7.2 7.2 7.2 7.2 5.6c 5.3c 5.4d 4.2c 4.8c 5.6d 5.1a 5.1a 4.0a 5.0a 5.0a 4.5a 4.1a 4.2c 4.8c 5.2d 4.5 5.4 6.1 5.5 5.5 5.5

805.88 200.13 24.72 71.4 100.7 40.3 29.8 18.07 130.34 60.32 17.15 25.26 48.24 22.22 57.5 34.66 22.3 50.5 20.17 18.75 18 55.18 29.56 103.82 17 63.16 96.69 27.47 17.55 11.54 71.5 48.6 20 20.2 7.79 57.03 18.7 12.33 21.25 12.99 10.68 23 14.2 16.3 24 7.97 19.01 10.67 17.26 9.15 30.3 18.13 40.25 30.42 15.63 9.25

28.3 109.1 35.5 124 46.9 50.8 42.9 28 51.5 21.13

13.5 175.95 17.7 17.7 21.0 47.1 8.5 60.7 17.7 8.4 22.9 19.7 12.0 28.0 17.8 50.9 14.2 14.3 10.9 7.5 28.2 41.2 59.11 27.99 26.23 57.06 27.99 44.8 46.75 23.1 65.1 23.59 20.45 24.73 35.44 37.56 59.27 27.56 62.46 79.01 20.84 23.29

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 Table 1 (continued) No. Station Code Station coordinates Lat. 200 Iznik Karayollari Sefligi 201 Denizli Bayindirlik ve Iskan Mud. 202 Burdur Bayindirlik ve Iskan Mu d. 203 Erzurum Bayindirlik ve Isk. Mu d. 204 Erzurum Bayindirlik ve Isk. Mu d. 205 Bolu Bayindirlik ve Iskan Mu d. 206 Van Bayindirlik ve Iskan Mu d. 207 Afyon Bayindirlik ve Iskan Mu d. 208 Afyon Bayindirlik ve Iskan Mu d. 209 Burdur Bayindirlik ve Iskan Mu d. 210 Andirin Tufan Pasa Ilko gretim O. 211 Akyazi Orman Isl. 212 Bornova Ziraat Fak. Dek. 213 Bingo l Bayindirlik ve Iskan Mud. 214 Bingo l Bayindirlik ve Iskan Mud. 215 Du zce Meteoroloji Istasyonu 216 Canakkale Meteoroloji Ist. 217 Sarayko y Jeotermal Lojmanlari 218 Denizli Bayindirlik ve Isk. Mud. 219 Sarayko y Jeotermal Lojmanlari 220 Sarayko y Jeotermal Lojmanlari 221 Denizli Bayindirlik ve Isk. Mud.
a

277

Earthquake Epicenter (date) coordinates Lat. 23.08.2000 40.68 04.10.2000 37.91 02.02.2001 37.64 29.05.2001 39.85 Lon. 30.71 29.04 30.19 41.55

PGA (gal) Re (km) NS 85.3 12.7 7.6 25.2 31.7 21.69 49.13 21.15 21.88 19.53 EW 16.21 66.38 30.12 17.21 21.94 131.64 24.81 94 51.5 21.3 50.42 23.13 37.11 74.62 20.23 31.86 15.56 123.23 45.84 34.46 UD 8.25 49.32 24.02 15.14

Site Mw C.

Lon. 29.75 29.114 30.221 41.262 41.262 31.61 43.406 30.561 30.561 30.221 36.34 30.622 27.229 40.503 40.503 31.149 26.4 28.923 29.114 28.923 28.923 29.114

IZN DNZ BRD ERZ ERZ BOL VAN AFY AFY BRD

40.44 37.813 37.704 39.903 39.903 40.747 38.504 38.792 38.792 37.704

3 2 2 1

5.5 5.0 4.6c 4.9 5.2 5.2 4.8 6.6 5.9 4.4d 4.8 4.3d 5.7 4.7d 4.2d 4.4 5.8 5.4 5.4 4.9 5.5 5.5

10.07.2001 39.8273 41.62

26.703 1 44.06 33.78 35.5 28 31.25 32.23 11.63 17.36 35.86 14.16 16.91 9.12 60.68 19.99 36.25 3 2 3 3 2 1 3 3 2 2 3 3 3 2 3 3 2

26.08.2001 40.9455 31.5728 22.3 189.07 02.12.2001 38.617 03.02.2002 38.581 03.02.2002 38.685 43.294 31.248 30.835 15.9 29.85

64.0 113.5 26.5 40.5 28.93 76.87

03.04.2002 37.8128 30.2572 12.5 14.12.2002 37.472 36.221 15.9

AND 37.58 40.67 38.455 38.897 38.897 40.844 40.142 37.932 37.813 37.932 37.932 37.813

09.03.2003 40.7328 30.6205 7.0 19.17 10.04.2003 38.2568 26.8345 40.8 78.58 01.05.2003 38.9737 40.534 8.9 152.25 01.05.2003 38.9613 40.341 21.05.2003 40.87 30.98 15.7 14.5 19.62 17.82 26.18 90.16 22.19 47.54

06.07.2003 40.42 26.21 34.8 23.07.2003 38.1718 28.8533 27.3 23.07.2003 38.1718 28.8533 45.9 26.07.2003 38.11 26.07.2003 38.11 26.072003 38.11 28.88 28.89 28.89 20.1

20.0 107.51 38.4 23.74

121.12 153.97 25.79 21.67

Converted from Ms. Converted from ML. c Converted from Mb. d Converted from Md.
b

the purpose, the site conditions reported in the databases of several institutions and the literature mentioned above are carefully compared, and the site conditions, which were same in more than one database reported by the institutions, were selected. For

the rest of the stations, for which different site conditions are defined by various institutions, the information from ERD (2003), Gu lkan and Kalkan (2002), and Bard (2002), Durukal et al. (1998) and Zare ISESD (2003) were chosen. In addition, site condition

278

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

at Bingo l station, which was observed by one of the authors of this recent study (Aydan et al., 2003), was also considered. Some recent attenuation models distinguish between the ground motion from reverse and strike slip earthquakes with the ground motion from reverse earthquakes being larger than for strike slip earthquakes. Due to the small number of normal faulting earthquakes in most strong motion data sets, the difference between ground motions for strike slip and normal faulting earthquakes has not been included in most attenuation relations (Douglas, 2003). The observations performed in the last decade suggested that even the hypocentral distance is the same; the acceleration values may be different depending on the place of the strong ground motion station with respect to the causative fault and its hanging wall or footwall (Abrahamson and Somerville, 1996; after Aydan and Hasgu r, 1997). Particularly accelerations recorded at the stations founded on the hanging wall may be greater than those obtained from the stations on the footwall. Although effect of faulting type can be important, Aydan and Hasgu r (1997), who assessed recorded acceleration values with fault type for some Turkish earthquakes, indicated that type of faulting seems to have less influence on the observed PGA

values. These investigators, however, suggested that this aspect should be further checked in the light of available data in the future. Gu lkan and Kalkan (2002), who more recently examined the peak ground motion data from the small number of normal faulting and reverse faulting earthquakes in the data set of Turkey, indicated that they were not significantly different from ground motion characteristics of strike slip characteristics, and normal, reverse or strike slip earthquakes could be combined into a single fault category. It is also noted that focal plane solutions of the most earthquakes selected from the existing database for this recent study are not available. Therefore, in this study, type of faulting was not considered in the development of the attenuation relationship. Among all the three-component accessible records, 221 records were selected from 122 earthquakes occurred between 1976 and November 2003 following the data selection criteria outlined above (PGA z 20 gal, Mw z 4, and 5 km V Re V 100 km). The data set employed in this study is given in Table 1 and distribution of all records with respect to magnitude, distance to epicenter and site condition is shown in Fig. 4. Station names in Table 1 are given in Turkish and are not translated. Based on this set of data, magnitudes of the earthquakes ranged between 4.1 and 7.5 (Mw).

Fig. 4. The distribution of records in the database employed in this study in terms of magnitude, distance to epicenter and site conditions.

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

279

3. Attenuation relationship development and comparison with some selected relationships for PGA In the development of the attenuation relationship, moment magnitude (Mw), distance to epicenter (Re), site condition (SA and SB, where SA = SB = 0 for rock sites, SA = 1 and SB = 0 for soil sites, and SA = 0 and SB = 1 for soft soil sites) and recorded largest horizontal PGA value of each station were employed. A total of 55 rock sites, 94 soil sites and 72 soft soil sites were considered in the analyses. In the first stage of the analyses, one coefficient for each of these terms was found via multiple regressions, and an attenuation relationship was derived for PGA. Then the PGA values predicted from this relationship and the observed PGA values in the database were subjected to non-linear regression to obtain the final attenuation relationship. Based on the analyses, the following attenuation relationship was obtained. PGA 2:18e
0:021833:3Mw Re 7:8427SA 18:9282SB

Fig. 6. PGA values predicted from Eq. (6) versus observed PGA values (r: correlation coefficient; S.D.: standard deviation; n: number of data).

The variation of PGA with distance to epicenter for rock, soil and soft soil sites with respect to different values of Mw is shown in Fig. 5. The general performance of the attenuation equation developed in this study is shown in Fig. 6, where

Fig. 5. Curves of PGA versus distance to epicenter for various magnitudes and different site conditions.

measured PGA values from the database are plotted against predicted PGA values using Eq. (6). As seen from Fig. 6, although a few points fall above and below the lines with 1:0.5 and 1:2 slopes, which indicate some overestimates and underestimates, respectively, most of the predictions are scattered within these lines. Particularly smaller PGA values fall close to the line 1:1. Correlation coefficient and standard deviation corresponding to PGAobserved = PGApredicted condition are 0.63 and 86.4, respectively. The PGA values predicted from Eq. (6) were also compared to those predicted from the domestic and some imported attenuation relationships. For the purpose, three domestic equations based on Turkish database and developed by Inan et al. (1996), Aydan (2001) and Gu lkan and Kalkan (2002) were employed. In the selection of imported equations for comparison, the following criteria were considered: (i) range of magnitudes employed in derivation of the imported equations should be similar those employed in this study, (ii) distance measures employed should be similar to that used in this study or should be available from the reports of various institutions, and (iii) the database employed in derivation of the imported equations should be collected from the region with tectonic regimes similar to that in Turkey. Two imported equations generally satisfying the above-

280

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

mentioned criteria were found in literature. These were developed by Theodulidis and Papzachos (1992) and Ambraseys et al. (1996). Theodulidis and Papzachos (1992) used a database consisting of 105 horizontal components from shallow earthquakes in Greece of magnitude 4.5 to 7.0, and of 16 horizontal components from four shallow earthquakes in Japan and Alaska of magnitudes 7.2 to 7.5. These investigators considered Ms and Mw, and distance to epicenter as magnitude scales and distance measure, respectively, and used two site categories such as alluvium and rock sites. Ambraseys et al. (1996) used 422 dada from shallow earthquakes in Europe, Middle East and Turkey. They considered distance to projection of rupture plane on surface (Rcl) for the earthquakes with Ms > 6, otherwise Re as a distance measure. They used four site conditions as rock (1), stiff soil (2), soft soil (3) and very soft soil (4), but considered only three because only three records from very soft soil were available. In order to compare Eq. (6) to the abovementioned five attenuation relationships main shocks of five great and one moderate earthquakes of Turkey (1992 Erzincan, 1995 Dinar, 1998 Adana-Ceyhan, 1999 earthquakes) Kocaeli, 1999 Du zce and 2000 Akyazy with Mw > 6, and a total of three aftershocks from the 1998 Adana-Ceyhan and 1999 Kocaeli earthquakes with Mw > 5 were selected from Table 1. During the comparisons the following considerations were made: (a) Because Aydans equation uses distance to hypocenter and he preferred to employ focal depths from ETHZ and if such parameters are not available, then he utilizes the parameters determined by USGS or Harvard (Aydan, 2003), in computations the depths reported by these institutions were employed. By considering that Ambraseys et al. (1996) and Gu lkan and Kalkan (2002) employ Rcl as a distance measure, among the selected earthquakes mentioned above, only the earthquakes with known Rcl values, which have been reported by the institutions were taken into comparison. However, for some of smaller events, rupture surfaces have not been defined clearly, therefore, distances to epicenter were also used for Gu lkan and Kalkans equation when Rcl is not available. (b) Site condition categories 3 and 4 used by Ambraseys et al. (1996) for soft and very soft

soils were combined into a single group as soft soil. Because the equations developed by Aydan (2001) and Theodulidis and Papzachos (1992) use only two site conditions (rock and soil), in

Fig. 7. Comparison of data from the 1996 Erzincan (a) and 1995 Dinar (b) earthquakes with the curves of the attenuation relationships at soil sites (h: focal depth; Re: distance to epicenter; Rcl: closest horizontal distance between the recording station and a point on the horizontal projection of the rupture zone on the earths surface).

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

281

Fig. 8. Comparison of data from the 1998 Adana-Ceyhan earthquake at soil (a) and soft soil (b) sites, and the aftershock of this earthquake at soft soil site (c) with the curves of the attenuation relationships.

Fig. 9. Comparison of data from the 1999 Kocaeli earthquake with the curves of attenuation relationships at rock (a), soil (b) and soft soil (c) sites.

282

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Fig. 10. Comparison of data from two aftershocks (a and b) of the 1999 Kocaeli earthquake with the curves of the attenuation relationships at rock, soil and soft soil sites.

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

283

this study, both equations were used for the stations founded on both soil and soft soil sites. (c) Although type of the magnitude for the equation developed by Inan et al. (1996) has not been mentioned in literature, these investigators are from ERD, which generally reports the earthquakes in ML and Md scales. Therefore, one of these two magnitude scales available was considered to predict the PGA values from Inan et al.s equation. (d) The comparisons for each earthquake with respect to site conditions were made separately. However, these equations employed different definitions for source to site distance. By considering these differences, x-axis was called source to site distance (km), and the measured points are shown by different symbols for Re and Rcl. In other words, the performance of the equations developed by Ambraseys et al. (1996) and Gu lkan and Kalkan (2002) was evaluated using Rcl versus PGA on the plots, which also show the curves of the other relationships using Re. The attenuation of PGA for the selected Turkish earthquakes for rock, soil and soft soil sites are compared in Figs. 7 12. The observed database points from these events were also shown on these

Fig. 11. Comparison of data from the 1999 Du zce earthquake with the curves of the attenuation relationships at rock (a), soil (b) and soft soil (c) sites.

earthquake with Fig. 12. Comparison of data from the 2000 Akyazy the curves of the attenuation relationships at soft soil.

284

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

curves to illustrate how well they fit the predictions. The comparisons indicate that although the attenuation relationships suggested by Inan et al. (1996) and Theodulidis and Papzachos (1992) show better fits to a few observed data, they generally considerably overestimates the PGA values for different site conditions. This overestimation becomes more evident particularly at the near source areas and for big earthquakes. The second attenuation relationship, which was used for comparison and developed by Ambraseys et al. (1996), generally underestimates the peak acceleration values particularly for some soil and soft soil sites (Figs. 8b,c, 10bsoil, 11c), while slightly overestimates for rock sites (Figs. 9a and 11a). Comparison of the other three relationships developed using Turkish database (Aydan, 2001; Gu lkan and Kalkan, 2002; this study) suggests that although many data points fall very close to the curves representing these relationships, they also yield some underestimations and overestimations. Among the equations considered for comparison, the equations developed in this study and by Aydan (2001) generally yield similar estimates, which are closer to the observed PGA values, particularly for rock sites (Figs. 9a, 10b and 11a). It is also noted that for soft soil sites during Du zce earthquake Aydans equation yields a good estimation for Bolu station, while the estimation from the equation developed in this study is very close to that recorded at Du zce station. In addition, for this earthquake all equations compared in this study overestimate the PGA values for soil sites. Aydan (2003) indicates that the estimations of the PGA for this earthquake probably deserve more detailed and sophisticated functions by considering not only rupture propagation and ground conditions, but also the topographical effects.

4. Iso-acceleration map of Turkey A probabilistic seismic hazard map of Turkey constructed by Gu lkan et al. (1993) replaced seismic zones in Turkey, and a total of 72 different combinations were used in Bayesian sense to derive a weighted average map corresponding to four different periods. Then, the seismic hazard zonation map based on this probabilistic approach was published by the

Ministry of Public Works and Settlement of Turkey (1996). Based on this map, Turkey is divided into five subclasses of seismic zone with PGA values of >0.4g, 0.3 0.4 g , 0.2 0.3 g , 0.1 0.2 g and < 0.1 g are assigned for zones ranging from I to V, respectively. The current practice in Turkey is to directly use the PGA values from the seismic codes published by the Ministry of Public Works and Settlement of Turkey (1998). According to the codes, PGA values of 0.4g, 0.3g, 0.2g and 0.1g are assigned for zones ranging from I to IV, respectively. The seismic zoning map of Turkey has some setbacks and these were outlined by Kayabali and Akin (2002). In order to minimize the effects of the setbacks in the present map, new seismic hazard maps were constructed by Kayabali (2002) and Kayabali and Akin (2002) using the probabilistic and deterministic approaches, respectively. However, Kayabali and Akin (2002) indicate that the probabilistic-based maps appear to have employed relatively large seismic zones, and therefore, deterministic-based seismic map seems to be more realistic. But in the construction of the iso-acceleration and seismic hazard maps of Turkey by Kayabali and Akin (2002), they did not consider fault segmentation concept and connected the segments of the main faults of Turkey in their assessments. In addition, they assumed that a fault would create a surface rupture equivalent to 1/3 of its total length and yield the maximum magnitudes. These assumptions resulted in very long faults, and consequently very high magnitudes were obtained when compared to those of the earthquakes associated with these faults. In this previous study, location of the epicenters particularly those falling in the areas, where active faults are not known, were not considered. Based on the magnitudes estimated, they employed the attenuation relationship proposed by Sadigh et al. (1997), who did not account data from normal faulting, for the construction of an iso-acceleration and seismic hazard maps of Turkey. In addition, the PGA values picked up from the iso-acceleration map constructed by these investigators are for bedrock and it is necessary to carry out site response analyses for the sites underlain by soil to compute the maximum PGA. A deterministic based iso-acceleration map of Turkey accounting different site conditions and based on the PGA values derived from the attenuation relationship developed in this study was also constructed. For

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

285

the purpose, the active fault map of Turkey by Saroglu et al. (1992) was employed as the main material to define the earthquake sources. In addition, based on the more recent studies on some active faults, which have not been shown on the map by Saroglu et al. (1992), were also included in the present assessments. A completed form of the active fault map employed in this study is shown in Fig. 13a together with the numbers assigned to the faults. The segments of the main faults were not connected and each segment was separately evaluated. Thus a total of 92 main faults (a total of 141 individual faults) were considered in the model. Name, length and type of each fault with the references related to these faults are listed in Table 2. It is apparent from Fig. 13a, that there are some regions free from active faults. However, Fig. 13b indicates the epicenters of a number of earthquakes, which occurred between 1900 and November 2003, appear in these regions. This situation suggest that only the use of distances to the known active faults in the attenuation relationship will result in unrealistic PGA assignments for a series of points selected in such regions. Therefore, in this recent study, the

epicenters were also decided to be used as the second group of earthquake source. For fault sources, the magnitude of the upper level earthquake is usually estimated from fault dimensions. Before fault segmentation concepts were developed, usually some fraction of the total fault length was used to estimate the magnitude of the design earthquake. For example, it was common to use 1/3 to 1/2 of the total fault length for the estimation of maximum magnitudes (Mark, 1977). Fault segmentation studies have replaced this approach for well-studied faults (Abrahamson, 2000). Therefore, in this study, use of fault segments is considered to be more realistic in the prediction of magnitudes, instead of connecting the segments. For a specific fault, the moment magnitude of the potential earthquake can be estimated by relating it to the potential rupture length of the fault using the relation proposed by Wells and Coppersmith (1994). However, it was considered that use of a relationship between magnitude and surface rupture length based on Turkish earthquake data would be more realistic. Therefore, the relation proposed by Aydan (1997) was preferred. This relation is given in

Fig. 13. (a) Active faults compiled from several investigators to be used in this study, and (b) distribution of the epicenters of the earthquakes occurred in Turkey between 1900 and March 2003.

286

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 Table 2 (continued) Fault Fault name number (segment no. in Fig. 7a) 20 21 South-East Anatolian thrust Semdinli Yu ksekova fault zone (21-1) (21-2) NA (22-1) Karacadag fault (22-2) NA (23-1) NA (23-2) Bozova fault Tut fault Tutak fault (26-1) Ercis fault (26-2) Caldiran fault Hasantimur lake fault Igdir fault (29-1) (29-1) Balik Go lu fault zone (30-1) (30-2) (30-3) Dogubayazit fault Kagizman fault Malazgirt fault Su phan fault Erzurum fault zone (35-1) (35-2) (35-3) (35-4) (35-5) (35-6) (35-7) (35-8) Karatas Osmaniye fault zone (36-1) (36-2) su O n fault Mut fault zone Karadag fault Sultandagi fault Burdur fault Golhisar-Cameli fault zone (42-1) (42-2) (42-3) (42-4) Dinar graben (43-1) (43-2) Length Fault Maximum Referenceb (km) typea magnitude (Mw) assigned 744.4 T 51.8 SS 47.1 SS 80.2 SS 22.2 U 18.3 42.9 51.6 25.1 108.0 33.7 55.7 15.6 67.9 30.0 108.7 41.3 39.1 50.8 90.1 21.3 47.0 168.6 169.5 29.7 42.7 57.0 43.4 44.0 40.4 59.4 29.9 27.1 42.6 23.6 145.8 59.1 81.7 32.0 25.4 18.1 17.8 19.9 SS N SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS U U U N N N N N N N N 7.7 6.5 6.4 6.8 5.8 5.7 6.3 6.5 5.9 7.0 6.1 6.5 5.6 6.7 6.1 7.0 6.3 6.3 6.5 6.9 5.8 6.4 7.4 7.4 6.1 6.3 6.5 6.3 6.3 6.3 6.6 6.1 6.0 6.3 5.9 7.3 6.6 6.8 6.1 5.9 5.7 5.7 5.8 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

Table 2 List of active faults appearing in Fig. 13a and the information associated with the faults considered in the construction of isoacceleration map Fault Fault name number (segment no. in Fig. 7a) 1 NAFZ: Marmara segment (1-1) (1-2) (1-3) NAFZ: Etili Sariko y segment (2-1) (2-2) (2-3) NAFZ: Yenice Go nen Bursa segment (3-1) (3-2) NAFZ: Du zce segment NAFZ: (5-1) (5-2) (5-3) (5-4) (5-5) Ezinepazari fault Akpinar Kirsehir fault zone (7-1) (7-2) Tuzgo lu fault zone Gumuskent fault Ecemis fault zone (10-1) (10-2) (10-3) (10-4) (10-5) Deliler fault Malatya Ovacik fault zone (12-1) (12-2) (12-3) Elbistan fault Su rgu fault EAFZ (15-1) (15-2) (15-3) Kavakbasi fault (South of Mus) Mus thrust Bingo l Karakocan fault Genc fault Length Fault Maximum Referenceb (km) typea magnitude (Mw) assigned 192.3 SS 192.5 SS 60.9 SS 60.5 SS 56.1 SS 237.7 SS 73.6 SS 7.5 7.5 6.6 6.6 6.5 7.6 6.7 2 2 2 1 1 1 1

22

23 24 25 26 27 28 29 30

4 5

108.5 SS 66.7 SS 169.6 98.6 107.0 310.5 154.7 262.5 40.3 23.6 187.3 94.3 97.7 128.2 102.2 55.3 90.1 77.2 107.5 71.4 64.0 67.9 62.3 269.4 207.4 122.7 86.5 SS SS SS SS SS SS SS SS N SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS SS

7.0 6.7 7.4 7.0 7.0 7.8 7.3 7.7 6.3 5.9 7.4 6.9 7.0 7.2 7.0 6.5 6.9 6.8 7.0 6.7 6.6 6.7 6.6 7.7 7.5 7.1 6.9 6.9 6.5 6.0

1 1 1 1 1 1 1 1 3 3 4 1 4 4 4 4 4 1 1 1 1 1 1 1 1 1 1 1 1

6 7

8 9 10

31 32 33 34 35

11 12

36

13 14 15

37 38 39 40 41 42

16 17 18 19

88.1 T 51.7 SS 26.4 SS

43 1

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 Table 2 (continued) Fault Fault name number (segment no. in Fig. 7a) 44 Eskisehir Sultanhani fault system (44-1) (44-2) (44-3) (44-4) Kaymaz fault Eskisehir fault zone Inonu Dodurga fault zone (47-1) (47-2) Kutahya fault Simav fault Akhisar fault Zeytindag Bergama fault zone Aliaga fault Demirkopru fault (53-1) (53-2) Gediz graben (54-1) (54-2) Bu yu k Menderes graben (55-1) (55-2) Denizli Honaz fault Karaova Milas fault (57-1) (57-2) Mugla Yatagan fault ren fault zone Ula O Sandikli fault (60-1) (60-2) Dazkiri .Cardak fault Kas fault Marmaris Ko ycegiz fault (63-1) (63-2) Kumdanli fault Beysehirgolu fault Gediz Dumlupinar fault (66-1) (66-2) Sancak Uzunpinar fault Merzifon fault Dodurga fault Derinkuyu fault Fethiye fault Bala fault Edremit fault NA Length Fault Maximum Reference (km) typea magnitude (Mw) assigned 76.3 N 58.2 77.9 40.4 33.1 43.3 77.2 38.5 40.3 56.6 35.3 61.8 N N N N N N N N N N N 6.8 6.6 6.8 6.3 6.1 6.3 6.8 6.2 6.3 6.5 6.2 6.6 5.9 5.6 5.5 7.2 7.4 7.2 7.5 6.6 6.4 6.2 6.4 6.6 6.0 5.8 6.3 5.7 5.7 5.7 6.4 6.0 6.1 6.4 6.5 6.2 5.7 5.7 5.5 5.9 6.5 6.4 5 5 5 5 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
b

287

Table 2 (continued) Fault Fault name number (segment no. in Fig. 7a) 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 Length Fault Maximum Referenceb (km) typea magnitude (Mw) assigned 7.3 6.7 6.7 6.3 5.3 5.5 5.9 7.0 5.7 6.1 5.6 5.6 6.6 6.8 6.7 5.7 6.4 6.4 6.2 1 1 1 5 1 1 1 6 1 1 1 1 4 3 3 3 1 1 1

45 46 47

48 49 50 51 52 53

24.1 N 16.7 N 14.1 130.2 168.9 134.7 N N N N

54 55

Cildir lake fault 157.1 SS Erivan fault 65.7 SS Karsanti Karaisali 65.6 SS fault zone Altinekin fault 41.5 N NA 11.2 U NA 13.9 U NA 24.0 SS Trakya fault 102.2 SS NA 17.5 U NA 31.9 U NA 16.5 U NA 17.0 U Sorgun fault 64.8 SS Sarikaya 77.4 SS Akdagmadeni fault Delice (Yerkoy) fault 65.9 SS Kirsehir fault 18.8 SS Tatarli fault 48.2 N Acigo l fault (92-1) 46.8 N (93-2) 35.9 N

56 57

189.0 N 61.5 N 44.6 N 36.9 46.3 58.3 28.5 20.9 41.5 19.8 18.2 18.1 50.2 29.5 33.6 N N N N N N N N N N N N

58 59 60 61 62 63

NAFZ: North Anatolian Fault Zone; EAFZ: East Anatolian Fault Zone. a SS: Strike slip fault; N: Normal fault; T: Thrust or reverse fault; U: Although shown on the active fault map of Turkey (Saroglu et al., 1992), information on fault type is not available. NA: Fault name is not assigned on the available maps. b 1: Saroglu et al. (1992); 2: Stein et al. (1997); 3:Dirik (1998); 4: Dirik and Goncuoglu (1996); 5: Dirik and Erol (2003); 6: Yaltirak et al. (1998).

64 65 66

67 68 69 70 71 72 73 74

46.1 N 50.4 SS 36.4 19.7 18.2 13.4 22.7 52.3 45.2 SS SS N U N N U

two forms, as fault rupture length versus Ms (Aydan, 1997) and fault rupture length in logarithmic scale versus Ms (Aydan et al., 2002). In this study, the latter relation was employed (Fig. 14). Assuming that each segment of a main fault and an individual fault would create a surface rupture equal to its total length measured from the map shown in Fig. 13a the maximum magnitudes (Ms) were estimated from Fig. 14 and, then converted to Mw using the equations found from Fig. 2a (Table 2). However, only the Bitlis suture zone extending throughout southeastern Anatolia (Fig. 13a) was assumed to cause a surface rupture equivalent to 1/3 of its total length, when it is broken (Table 2). For the construction of the iso-accelaration map of Turkey, a computer program was written in BASIC

288

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

Fig. 14. Relation between surface magnitude (Ms) and surface rupture length (L) based on the Turkish earthquakes (after Aydan et al., 2002).

programming language. All faults and earthquake epicenters were identified as a series of points and as single points with global coordinates, respectively. Then, the area between the coordinates of 25.5 45.5j eastern longitudes and 35.5 42.5j northern latitudes was divided by 0.1j intervals and a grid system composed of 14271 points was obtained using the

computer program. The program calculated closest distances to all faults shown in Fig. 13a and to the epicenters (Fig 13b) for each grid point. For the purpose, the points defining each fault zone were considered and the two closest points on each fault to the grid point under consideration were found. Then the shortest distance between the grid point and the fault was calculated. The distances from the grids to the epicenters were also computed by the program. The maximum magnitudes assigned to each fault and/or segment (Table 2) and obtained from the database for each epicenter were separately employed in the attenuation relationship (Eq. (6)) suggested in this study to predict the PGA for each point under consideration. Then two iso-acceleration maps based on these two approaches (fault segments and epicenters) were constructed. Finally, both maps were compared by pixel to pixel and the highest of these PGA values was assigned to each point. Fig. 15 shows the final iso-accelaration map of Turkey constructed by following the steps given above. It should be remembered that the PGA values picked up from this map is only for rock sites. In order to estimate the values of PGA for soil and soft soil sites, the picked up value from the map should be multiplied by 1.186 and 1.511, respectively. As can be seen from Fig. 15, higher values of PGA are generally concentrated along the main seismotec-

Fig. 15. Base iso-acceleration map of Turkey constructed using the attenuation relationship suggested in this study (PGA contours are in gal and represent rock sites; for soil and soft soil sites multiple the PGA value picked up from the map by 1.186 and 1.511, respectively).

R. Ulusay et al. / Engineering Geology 74 (2004) 265291

289

tonic features of Turkey. However, in addition to the faults, the use of the distances to the epicenters, which has not been considered in construction of the previous iso-acceleration maps of Turkey, enabled a better consideration on the PGA values to represent some regions far from and/or free from the faults.

5. Conclusions In this study, an attenuation relationship of PGA based on the recent Turkish database was presented. In addition, an attempt was made to construct an isoaccelaration map using the proposed prediction model and deterministic approach for Turkey. The proposed attenuation relationship seems to be helpful for the prediction of PGA for earthquakes with magnitude (Mw) ranging between 4.1 and 7.5, and distance to epicenter equal and/or less than 100 km with respect to rock, soil and soft soil site conditions. The comparison between the attenuation relationship suggested in this study and some imported relations developed using data from Europe and Middle East indicated that the relationship of Theodulidis and Papzachos (1992) considerably overestimated the PGA values, while Ambraseys et al.s (1996) equation generally yields underestimated values particularly for soft soil sites. Similarly, the relationship based on Turkish data and developed by Inan et al. (1996) yields highly overestimated PGA values, especially at near source areas. Among the attenuation relationships used for comparison, the equations developed for Turkey by Aydan (1997, 2001), Gu lkan and Kalkan (2002) and the authors of this study yield better match with observed data. Therefore, it can be concluded that the use of attenuation relationships based on Turkeys own data should be preferred to predict more precise PGA values. The suggested iso-acceleration map, which is a base map considering only rocky ground condition, was constructed using the PGA values predicted from a model based on the data from Turkey and also offers an opportunity to estimate PGA values for soil and soft soil site conditions when the given coefficients for these site conditions are multiplied by the PGA values picked up from the map for a certain point. It is concluded that the map constructed can be considered

as a base map for a further modification of the previously suggested seismic hazard zonation map of Turkey. Although the attenuation relationship developed in this study and the other two domestic relationships (Aydan, 1997, 2001; Gu lkan and Kalkan, 2002) mentioned above predict more precise PGA values, it is recommended that in order to improve the precision of the ground motion estimates site characterization parameters based on shear-wave velocity measurements at each station, which are currently lacking in Turkey, need to be included into the equations. In addition, as recommended by Aydan and Hasgu r (1997) and Aydan (2003) directivity effect, faulting type and the effect of overhanging side on footwall should be considered for when sufficient data on these factors become available.

Acknowledgements The authors would like to express their sincere for thanks to Dr. John Douglas and Dr. Mehdi Zare their critical reviews and valuable comments that lead to significant improvements to the article. The authors mer Aydan (Tokai extend their gratitudes to Prof. Dr. O University, Japan), not only for his contributions based on his long-term studies on engineering and scientific aspects of Turkish earthquakes, but also for his valuable comments for this study. Thanks are also due to Assoc. Prof. Dr. Kadir Dirik of Hacettepe University for his valuable comments during the compilation of the active faults of Turkey. The authors are also grateful to the Earthquake Research Department of General Directorate for Disaster Affairs and Kandilli Observatory and Earthquake Research Institute of Bogazici University, and the other institutions, which provide online access to the Turkish strong motion and earthquake records of Turkey on the Internet.

References
Abrahamson, N.A., 2000. State of the practice of seismic hazard evaluation. GEOENG 2000 Conference Proceedings, Melbourne, Australia (on CD).

290

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 ERD (Earthquake Research Dept. of Turkey), 2003. http://www. deprem.gov.tr. Erdik, M., Doyuran, V., Gu lkan, P., Akkas, N., 1985. A probabilistic assessment of the seismic hazard in Turkey. Tectonophysics 117, 295 344. ETHZ, 2003. http://seismo.ethz.ch/. Gu lkan, P., Kalkan, E., 2002. Attenuation modeling of recent earthquakes in Turkey. J. Seismol. 6, 397 409. Gu lkan, P., Kocyigit, A., Yucemen, M.S., Doyuran, V., Basoz, N., 1993. Earthquake zoning map of Turkey based on most recent data. METU Eathquake Engineering Research Center, Report No. 93-01. 156 pp. (in Turkish). HARVARD, 2003. http://www.seismology.harvard.edu/CMTsearch. html. Inan, E., Colakoglu, Z., Koc, N., Bayu lke, N., Coruh, E., 1996. Earthquake Catalogs with Acceleration Records from 1976 to 1996 General Directorate of Disaster Affairs, Earthquake Research Department, Ankara, Turkey (98 pp., in Turkish). ISC, 2003. http://www.isc.ac.uk. ISESD, 2003. http://www.isesd.cv.ic.ac.uk/. Joyner, W.B., Boore, D.M., 1981. Peak horizontal acceleration and velocity from strong motion records including records from the 1979 Imperial Valley, California earthquake. Bull. Seismol. Soc. Am. 71, 2011 2038. Joyner, W.B., Boore, D.M., 1988. Measurement, characterization, and prediction of strong ground motion. Earthquake Engineering and Soil Dynamics: 2. Recent Advances in Ground Motion Evalution. ASCE, New York, pp. 43 102. Kayabali, K., 2002. Modeling of seismic hazard for Turkey using the recent neotectonic data. Eng. Geol. 63, 221 232. Kayabali, K., Akin, M., 2002. Seismic hazard map of Turkey using the deterministic approach. Eng. Geol. 69 (1/2), 127 137. Mark, R.K., 1977. Application of linear statistical model of earthquake magnitude versus fault length in estimating maximum expectable earthquakes. Geology 5, 464 466. Ministry of Public Works and Settlement of Turkey, 1996. Seismic Hazard Map of Turkey. Ministry of Public Works and Settlement of Turkey, 1998. Turkish Seismic Codes (85 pp., in Turkish). Sabetta, F., Pugliese, A., 1987. Attenuation of peak horizontal acceleration and velocity from Italian strong motion records. Bull. Seismol. Soc. Am. 77, 1491 1513. Sadigh, K., Chang, C.Y., Egan, J.A., Makdisi, F., Youngs, R.R., 1997. Attenuation relationships for shallow crustal earthquakes based on California strong motion data. Seismol. Res. Lett. 1 (68), 180 189. Saroglu, F., Emre, O., Kuscu, I., 1992. Active fault map of Turkey. Printed by General Directorate of Mineral Research and Exploration of Turkey. Stein, R.S., Barka, A.A., Dietrich, J.H., 1997. Progressive failure on the North Anatolian Fault since 1939 by earthquake stress triggering. Geophys. J. Int. 128, 594 604. Theodulidis, N.P., Papzachos, B.C., 1992. Dependence of strong ground motion on magnitudedistance, site geology and macro-seismic intensity for shallow earthquakes ib Greece: I, peak horizontal acceleration, velocity and displacement. Soil Dyn. Earthqu. Eng. 11 (7), 387 402.

Abrahamson, N.A., Somerville, P.G., 1996. Effects of the hangingwall and footwall on ground motions recorded during the Northridge earthquake. Bull. Seismol. Soc. Am. 86 (18), 593 599. Ambraseys, N.N., Simpson, K.A., Bommer, J.J., 1996. Prediction of horizontal response spectra in Europe. Earthquake Eng. Struct. Dyn. 25 (4), 371 400. ., 1997. The seismic characteristics and the occurrence Aydan, O pattern of Turkish earthquakes. Turkish Earthquake Foundation Report No.TDV/TR 97-007, 41 pp. ., 2001. Comparison of suitability of submerged tunnel Aydan, O and shield tunnel for subsea passage of Bosphorus. Jeol. Mu hendis. 25 (1), 1 17 (in Turkish). Aydan, O., 2003. Personnel Communication. Tokai University, Dept. of Marine Civil Engineering, Shizuoka, Japan. ., Hasgu Aydan, O r, Z., 1997. The characteristics of acceleration i waves of Turkish earthquakes. 4. Ulusal Deprem Mu hendislig i Tu Kongresi, Ankara, Deprem Mu hendislig rk Milli Komitesi, pp. 30 37, (in Turkish). ., Sezaki, M., Yarar, R., 1996. The seismic characteristics Aydan, O of Turkish earthquakes. 11th World Conference on Earthquake Engineering, Acapulco, Mexico. CD paper no. 1025. ., Kumsar, H., Ulusay, R., 2002. How to infer the possible Aydan, O mechanism and characteristics of earthquakes from the striations and ground surface traces of existing faults. JSCE, Earthqu. Struct. Eng. Div. 19 (2), 199 208. ., Ulusay, R., Miyajima, M., 2003. The Bingo Aydan, O l earthquake of May 1, 2003. Japan Society of Civil Engineers, Tokyo (93 p.). http://www.jsce.or.jp/report/report200307.pdf. Boore, D.M., Joyner, W.B., Fumal, T.E., 1997. Equations for estimating horizontal response spectra and peak acceleration from western North American earthquakes: a summary of recent work. Seismol. Res. Lett. 68 (1), 128 153. Campbell, K.W., 1981. Near-source attenuation of peak horizontal acceleration. Bull. Seismol. Soc. Am. 71 (6), 2039 2070. Campbell, K.W., 1997. Empirical near source attenuation relationships for horizontal and vertical components of peak ground acceleration, peak ground velocity, and pseudo-absolute acceleration response spectra. Seismol. Res. Lett. 68 (1), 154 179. COSMOS, 2003. http://www.cosmos.equ.org. Dirik, K., 1998. Neotectonic of Kirsehir and Vicinity. In: Kilic, R. (Ed.), Proc. of Panel on Seismicity of Kirsehir. Municipality of Kirsehir, Kirsehir, pp. 28 37 (in Turkish). Dirik, K., Erol, O., 2003. Tectonomorphologic evaluation of Tuz Go lu and surrounding area, Central Anatolia Turkey. Spec. Publ.-Bull. Turk. Petr. Geol. 5, 27 46. Dirik, K., Goncuoglu, M.C., 1996. Neotectonic characteristics of Central Anatolia. Int. Geol. Rev. 38, 807 817. Douglas, J., 2003. Earthquake ground motion estimation using strong-motion records: a review of equations for the estimation of peak ground acceleration and response spectral ordinates. Earth-Sci. Rev. 61 (1 2), 43 104. Durukal, E., Erdik, M., Avci, J., Yuzugullu, O., Alpay, Y., Avar, B., Zulfikar, C., Birol, T., Mert, A., 1998. Analysis of the strong motion data of the 1995 Dinar, Turkey earthquake. J. Soil Dyn. Earthqu. Eng. 17, 557 578.

R. Ulusay et al. / Engineering Geology 74 (2004) 265291 USGS, 2003. http://geopubs.wr.usgs.gov/open-file/of01-163/web/ sources_of_data.htm. Wang, B.-Q., Wu, F.T., Bian, Y.-J., 1999. Attenuation characteristics of peak acceleration in North China and comparison with those in the eastern part of North America. Acta Seismol. Sin. 12 (1), 26 34 on internet at: http://www.chinainfo.gov.cn/ periodical/dizhen-E/dzxb99/dzxb9901/990104.htm. Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area,

291

and surface displacement. Bull. Seismol. Soc. Am. 4 (84), 975 1002. Yaltirak, C., Alpar, B., Yu ce, H., 1998. Tectonic elements controlling the evolution of the Gulf of Saros (Northeastern Aegean Sea, Turkey). Tectonophysics 300, 227 248. , M., Bard, P.-Y., 2002. Strong motion dataset of Turkey: data Zare processing and site classification. J. Soil Dyn. Earthqu. Eng. 22, 703 718.