Ck. Saikia, MODELING OF STRONG GROUND MOTIONS FROM THE 16 SEPTEMBER 1978 TABAS, IRAN, EARTHQUAKE, Bulletin of the Seismological Society of America, 84(1), 1994, pp. 31-46
We have examined the adequacy of the published KS3 and L1 slip distrib
ution models developed for the M(S) = 7.4 16 September 1978 Tabas eart
hquake (Hartzell and Mendoza, 1991) by modeling the high-frequency acc
elerograms recorded at Dayhook, Boshrooyeh, and Tabas stations. The ag
reement in peak ground acceleration (PGA) and duration between data an
d simulated accelerograms was chosen as the criterion for the model ad
equacy. The fault was specified with a seismogenic rupture area 95 km
long and 45 km wide. We used a semi-empirical simulation method in whi
ch accelerograms from an aftershock from the 1979 Imperial Valley eart
hquake were used to represent the radiation pattern of P, SV, or SH wa
ves for a source-receiver geometry. The fault surface is divided into
many subfaults. The contributions from these subfaults, weighted by th
e slip amounts, were lagged and summed to simulate the accelerograms o
f the main event. We computed accelerograms assuming constant rake and
variable rake on the fault plane. The former was simulated using the
SUM slip model (i.e., the slip model representing the vector sum of th
e slip components at every point on the fault), and the latter was sim
ulated by summing the accelerograms generated by the slip models of th
e dip-slip and strike-slip faults. The influence of the rise time of t
he main event was investigated. The KS3 asperity model produces result
s marginally better than the L1 asperity model. The duration is predic
ted, consistent with the data. The peak amplitudes of the simulated gr
ound motions remain within a factor of 1.5 and 3 of the recorded data
at Dayhook and Boshrooyeh, respectively, for the variable-rake and wit
hin a factor for 2.5 at both stations for the fixed-rake angle (KS3 mo
del). The large PGA recorded on the Tabas accelerograms could not be r
eproduced, which warrants additional study. A notable agreement was ob
tained in the response spectra at Dayhook. This study demonstrates tha
t accelerograms generated by an earthquake in one tectonic region can
be transported to another tectonic region so that they can be used to
represent the empirical radiation pattern of the subfaults in simulati
ng ground motions for a large event.