FAULT LOCATION AND RUPTURE KINEMATICS OF THE MAGNITUDE 6.8, 1992 ERZINCAN EARTHQUAKE, TURKEY, FROM STRONG GROUND MOTION AND REGIONAL RECORDS

Citation
P. Bernard et al., FAULT LOCATION AND RUPTURE KINEMATICS OF THE MAGNITUDE 6.8, 1992 ERZINCAN EARTHQUAKE, TURKEY, FROM STRONG GROUND MOTION AND REGIONAL RECORDS, Bulletin of the Seismological Society of America, 87(5), 1997, pp. 1230-1243
Citations number
30
Categorie Soggetti
Geochemitry & Geophysics
ISSN journal
00371106
Volume
87
Issue
5
Year of publication
1997
Pages
1230 - 1243
Database
ISI
SICI code
0037-1106(1997)87:5<1230:FLARKO>2.0.ZU;2-U
Abstract
The hypocenter location of the 13 March 1992, M = 6.9 Erzincan earthqu ake (Turkey) and its rupture kinematics have been studied using local and regional seismic records to understand its relationship with the t ectonic activity of the Erzincan basin and of the North Anatolian faul t (NAF). Regional records of the mainshock and the largest aftershocks were used to locate the respective epicenters. The mainshock S-P trav el time delay and the P-wave polarization of the near-source accelerog ram record in Erzincan (ERZ), calibrated by the aftershock records, co nstrain the mainshock hypocenter to be located about 10 km to the east -southeast of Erzincan, 39 degrees 42.3 N, 39 degrees 35.2 E, and at a best-fit depth of about 9 km. The activated fault is a major branch o f the NAF along the eastern edge of the basin. Hypocentral depths grea ter than about 16 km are not supported by the data. The ground velocit y record at ERZ has been synthesized for constraining the northwestwar d rupture, assuming a kinematic model with constant slip and rupture v elocity, and a strike 125 degrees, for a large number of models. The l atter were constructed by varying the rupture parameters in a discrete domain. The best-fit fault models are located within 1 km of the epic enter above and have all their northwestern end within 5 km of ERZ. Th e up-dip limit of the rupture in the vicinity of ERZ ranges from 5 to 2 km in depth, in agreement with the absence of surface breakage. The estimated rupture velocity is 3 km/sec; the rise time is 0.35 sec; and the mean slip of about 1 m is poorly resolved, with a standard uncert ainty of 0.5 m, being strongly sensitive to the rupture velocity and t o the fault location. Combining these results with the seismic moment and the source duration provided by other studies leads to a total rup ture length ranging between 20 and 35 km. The rupture thus propagated bilaterally on the NAF and was confined within the limits of the basin , stopping to the southeast near the intersection of the NAF with the Ovacik fault.