Open Access
Faulting process of the 1990 June 20 Iran earthquake from broadband records
Author(s) -
Campos J.,
Madariaga R.,
Nábělek J.,
Bukchin B. G.,
Deschamps A.
Publication year - 1994
Publication title -
geophysical journal international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.302
H-Index - 168
eISSN - 1365-246X
pISSN - 0956-540X
DOI - 10.1111/j.1365-246x.1994.tb04673.x
Subject(s) - seismology , geology , epicenter , inversion (geology) , geodesy , seismic moment , directivity , fault plane , source model , waveform , frequency band , fault (geology) , physics , telecommunications , tectonics , computational physics , computer science , bandwidth (computing) , radar , antenna (radio)
SUMMARY We study the rupture process of the Iranian earthquake of 1990 June 20 from broad‐band data. We redetermined its moment tensor and source mechanism from long‐period surface waves. Then we studied the P and SH broad‐band recordings from the GEOSCOPE and IRIS seismic networks. We inverted the body waves using a full waveform modelling. For the inversion we used the gradient method of Nábělek, mixed with a limited exploration of the parameter space. In order to control the resolution of inversion, we proceeded by a stepwise procedure increasing the model complexity with every new inversion. We used the reduction of cost functional as a criterion for the validity of the inverted model. We first inverted the body waves considering a single‐point source model. We found that body waves are dominated by a large energy release about 20 s from the onset of the signals. In the next step, fixing fault‐plane solution, we inverted for the directivity of the source. We found clear evidence of propagation towards the east of the epicentre. We determined a rupture velocity of ∼2.5km s −1 . The preceding source inversion reduced the variance of the residuals by about 30 per cent. Further improvement of the fit of the body‐wave signals was obtained using an extended line source with constant fault‐plane solution. We obtained a very good fit with an asymmetrical model. Initially rupture is bilateral but north‐western propagation stops after 10 s. Later rupture continues unilaterally in the south‐eastern direction with a rupture velocity of the order of 2.5 km s −1 . Finally an inversion was attempted with a line source with varying fault mechanism. The final solution obtained by this procedure reduces the variance of the body‐wave residuals by 60 per cent with respect to the variance of the observed body‐wave signals. Variations in fault‐plane solution along the fault are considered to be well resolved.