ELSEVIER Tectonophysics 292 (1998) 255–266 Source inversion of the October 1, 1995, Dinar earthquake . M s D 6:1/: a rupture model with implications for seismotectonics in SW Turkey Ali Pınar * Kandilli Observatory and Earthquake Research Institute, Bo˘ gazic ¸i University, 81220, C ¸ engelko ¨y, Istanbul, Turkey Received 28 May 1997; accepted 13 January 1998 Abstract An earthquake of M s D 6:1 devastated the town of Dinar (SW Turkey, population 35,000) on October 1, 1995, killing 90 people and destroying 30% of the town. The earthquake generated complex body-waveforms varying with azimuth at teleseismic distances. The method of complex body-waveform inversion developed by Kikuchi and Kanamori (1991) was used to infer a source process for the earthquake. Two subevents were necessary to explain the observed seismic records. The inversion result suggests that the Dinar earthquake initiated at the SE end of the Dinar fault with a subevent of seismic moment M o D 0:5 ð 10 18 N m. Six seconds later, the second subevent took place about 10 km to the northwest of the first subevent with a seismic moment a few times larger than the first. The CMT depths of the first and second subevents were found to be 10 and 15 km, respectively. Both subevents had a predominantly normal faulting mechanism with slip-vectors oriented NE–SW, showing good agreement with the velocity-vectors obtained from the recent SLR and GPS studies as well as with the regional stress orientation obtained from geological data. The main shock was preceded by foreshock activity concentrated at the SE end of the Dinar fault where the first subevent took place, while the aftershock activity was concentrated in the vicinity of the second subevent. The spatial distribution of the foreshock and the aftershock activities and the locations of the subevents suggest that the first subevent broke an asperity and the second subevent broke a barrier on the fault, following the definition by Aki (1984) of an asperity and barrier earthquake model. About 10 km of surface ruptures were associated with the earthquake while the estimates yield a rupture length of 25 km. The calculated source parameters of the subevents and their locations suggest that the surface ruptures were probably associated with the first subevent. The estimates also show that the rupture zones of the two subevents overlapped where the maximum vertical displacement was observed.  1998 Elsevier Science B.V. All rights reserved. Keywords: source inversion; complex earthquakes; SW Turkey; Dinar fault; normal faulting 1. Introduction The October 1, 1995, Dinar earthquake . M s D 6:1/ devastated Dinar, a town with a population of 35,000 in SW Turkey, killing 90, injuring 260 and Ł Fax: C90 (216) 308-5223; E-mail: pinara@hamlin.cc.boun.edu.tr leaving more than 25,000 people homeless. About 30% of the buildings in the town collapsed or were heavily damaged (Erdik et al., 1995; Gu ¨ndo˘ gdu et al., 1996; Demirta¸ s et al., 1996). The epicentre was located at 38.12ºN, 30.02ºE by the Kandilli Ob- servatory and Earthquake Research Institute (KO- ERI) network and at 38.09ºN, 30.15ºE by the USGS (Fig. 1). A strong-motion recorder was located a few 0040-1951/98/$19.00  1998 Elsevier Science B.V. All rights reserved. PII S0040-1951(98)00056-0