Haluk Eyidoğan

The seismotectonics of the Marmara region (Turkey): results from a microseismic experiment

The Marmara region is an active tectonic zone characterised by the transition between the dextral strike-slip regime of the North Anatolian Fault (NAF ) and the extension regime of the Aegean Sea. Strong historical earthquakes (M>7) and the presence of known seismic gaps imply a high level of seismic hazard. A synthesis of recent studies of active tectonics in the region is presented, including inland and underwater observations. The branching of the NAF is explained in terms of increasing influence of the extension. Historical information and instrumental seismicity are reinterpreted in order to have a critical appraisal of the existence of large seismic gaps in the central and eastern Marmara Sea. Focal mechanisms of strong earthquakes are used to obtain orientation and shape factor of the deviator of the stress tensor. The resulting tensor is in shear regime (s 2 vertical ) but close to extension (R=0.93) with s 1 oriented N145°0. A microseismic experiment with 48 stations distributed around the Marmara Sea was carried out in October‐December 1995. A total of 137 microearthquakes were located and 23 of those were selected to obtain focal mechanism solutions. The epicentral distribution indicates activity along the system of pull-apart basins north of the Marmara Sea. The segment between Marmara Sea and the Saros Bay, activated in 1912, and the Gulf of Izmit, site of the 1754 earthquake, are now silent. Seismic activity is very linear along the northern branch of the NAF, but it is more diVused on the Bursa and Iznik branches, southeast of the Marmara Sea. The stress tensor obtained from the focal mechanisms of the micro-earthquakes is compared to the one inverted from teleseismic data. The microseismic stress is compatible with a shear (intermediate) regime like the one obtained from strong earthquakes, though not as well constrained, one of the acceptable solutions having the same orientation but diVerent shape (R=0.5).

High-Frequency Ground Motion in the Erzincan Region, Turkey: Inferences from Small Earthquakes

This study has been supported by Istituto Nazionale di Geofisica e Vulcanologia, INGV, Internal Project: “Attenuazione e leggi di scala nei paesi dell’area Mediterranea” (internally funded). R. B. Herrmann’s participation was supported by INGV and by the Earthquake Engineering Research Centers Program of the National Science Foundation under Award Number EEC-9701785.

The İzmit (Kocaeli), Turkey Earthquake of 17 August 1999: Previous Seismicity, Aftershocks, and Seismotectonics

A field experiment carried out from 17 July to 2 November 1999 with a dense seismic network surrounding the Marmara Sea region and including the rupture zone of the Izmit (Kocaeli) destructive earthquake of 17 August 1999 ( M W 7.6, Harvard centroid moment tensor); (Delouis et al., 2002) permitted the collection of precise seismic data and the quantitative evaluation of the associated set of surface ruptures. A previous experiment conducted in 1995 had already given reliable information about the distribution of seismicity in space, focal mechanisms, and the regional stress regime. Aftershock distribution and clustering, together with the variations of slip amplitude along the fault, allow the identification of at least five segments. A maximum horizontal displacement of about 5 m is observed in Golcuk and east of Sapanca Lake. The rupture is extremely linear, but its complexity increases toward the western end, including the bifurcation. The stress field associated with the aftershocks is compared with the one obtained in 1995, showing the invariance of the smallest principal axis. Lack of definition of the other principal axes suggests a transition from shear to extension. A clear quiescence period of 18 days preceded the Izmit earthquake. A precursor, which occurred 100 sec before, had the same epicenter as the mainshock. Manuscript received 30 August 2000.

Analysis and interpretation of the aftershock sequence of the August 17, 1999, Izmit (Turkey) earthquake

A micro-seismic field experiment has been carried out in the Marmara Searegion. The analysis of the events before and after the August 17, 1999Izmit (Turkey) earthquake has been completed. 1446 events have beenwell located out of a total of 3165 recorded within the period from July15 to November 2, 1999. 67% of the aftershocks with magnitudegreater than 4 have occurred within the first 6 days after the main-shock. Earthquakes of the Izmit sequence are distributed in the first 15 km of theearth crust, but major events are located in between 5 km and 15 kmdepth. The seismicity pattern defines a rupture plane extending for about150 km in an E-W direction. The rupture is extremely linear butsegmented, and its complexity increases towards the western endmanifesting bifurcation. A stress analysis has been carried out both at thewestern end and all along the aftershock zone. 96 selected aftershocks,registered between August 21 and October 22, were chosen in order tocompute their focal mechanisms and obtain information about the stressregime after the Izmit earthquake. Strike-slip and normal faultingmechanisms are dominant. The numerous strike-slip mechanisms arecompatible with a dextral motion on an EW oriented vertical fault plane. The best stress tensor solution shows a regime in extension with awell-defined σ3 axis oriented approximately N35°.

Present-Day Seismicity and Seismotectonics of the Cilician Basin: Eastern Mediterranean Region of Turkey

The nature of the interaction between the Anatolian, Arabian, and Af- rican plates is an important problem of northeastern Mediterranean tectonics. The Dead Sea Fault and East Anatolian Fault Zone are the most prominent structural elements in this region. The Cilician Basin, which is located at the junction of these faults, is known to have a very active seismic history. However, instrumental obser- vations have been insufficient because of the lack of an adequate local seismic net- work. To monitor ongoing seismic activity and to analyze the seismotectonics of the Cilician region (36 to 38 N, 34 to 38 E), a digital seismic network of 18 stations with velocity transducers has been operated by the Earth and Marine Sciences Re- search Institute (EMSRI) of the Scientific and Technical Research Council of Turkey (TUBITAK) since 1993. No previous studies have been carried out in this region for crustal structure. An optimum 1D velocity model was obtained by using the VELEST algorithm for reliable determination of earthquake hypocenter locations. Approxi- mately 2500 earthquakes detected between 1993 and 2002 with local magnitude greater than 1.5 have been located by using this optimum velocity model. The most notable feature of the epicentral distribution is the presence of dense clusters of activity along local fault trends. The majority of hypocentral depths of the events range between 5 and 35 km. Single-fault plane solutions of reliable events were calculated in order to analyze the seismotectonic characteristics of the region. Results show that the Cilicia region constitutes a wide left lateral shearing zone that indicates a diffuse plate boundary between the African, Arabian and Anatolian Plates.

Site attenuation and source parameters on the North Anatolian Fault zone, eastern Turkey estimated from the aftershocks of 13 March 1992 Erzincan earthquake

Site attenuation and source characteristics of 45 aftershocks of the 13 March 1992 Erzincan, eastern Turkey earthquake have been determined from SH-wave spectra using a least-squares best-fit method. Although the most of the seismograms were recorded on the ophiolitic rock sites and the average regional attenuation correction of Q(f) = 35 f0.83 was applied, the high fall-off rates γ ranging between 3.0 and 4.5 beyond the corner frequencies of 8–13 Hz were observed on the SH spectra. The site attenuation value, κ, has been calculated from the slope of the high frequency part of the SH spectra. It was found that κ varied in the range of 0.0124–0.0364 s and the average was 0.0246 s with a standard deviation of 0.0047 s. The high fall-off rates of observed spectra have decreased considerably as a result of site attenuation corrections and converged to an average value of 2.3. We concluded that the high fall-off rates of SH wave spectra are mainly controlled by highly deformed Miocene ophiolitic formations which covers a wide area in the North Anatolian Fault zone in the vicinity of Erzincan region. Using the spectra that were corrected for regional and site attenuation, and assuming a Brunes source model; seismic moment, source radius, and stress-drops of the aftershocks were computed. We found that stress-drops for some of the aftershocks in Erzincan area have slightly decreased after removal of site attenuation over SH wave observed spectra.