The 2007 Kahak and 2010 Kazerun earthquakes: constrained non-negative least-squares linear finite fault inversion for slip distribution

Document Type : Research

Authors

1 M.Sc. graduate, Department of Earth Physics, Institute of Geophysics, University of Tehran, Tehran, Iran

2 Academic member

3 Department of Earth Physics, Institute of Geophysics, University of Tehran, Tehran, Iran

Abstract

In this study, two moderate earthquakes from two main seismotectonic provinces of Iran are chosen for investigating slip distribution using finite-fault modeling. The first earthquake is the 2007 June 18 Mw 5.5 Kahak earthquake which is sited in Central Iran seismotectonic province in the vicinity of Kahak district of Qom province near Tehran, the capital of Iran. The second one is the 2010 September 27 Mw 5.9 Kazerun earthquake situated in Zagros seismotectonic province, near Kazerun County in Fars Province. This research aims to find finite-fault modeling of the broadband three-component displacement waveforms of these earthquakes using a least-squares inversion method for the spatial and temporal slip distribution. Green’s functions are calculated using the frequency-wavenumber integration code (FKRPROG) developed by Saikia (1994), and the inversion algorithm used for acquiring synthetic data is based on a stabilized constrained non-negative least-squares method introduced by Hartzell and Heaton (1983). A great many inversions are implemented to obtain the optimal parameters used in the process, including rupture velocity and rise time. The rupture velocity of 2.6 km/s (0.75 Vs) and the rise time of 1.4 s are used for the first event, and 2.8 km/s (0.75 Vs) and 2.1 s are chosen for the second one. Results show the rupture with the peak slip of 8.6 cm and 14.3 cm, and the total seismic moment release of 1.59e+24 dyne-cm and 2.80e+25 dyne-cm for the Kahak and Kazerun earthquakes, respectively. Furthermore, due to the non-uniqueness of the inversion problem, a set of solutions is presented for both events. Among these models, the final solutions for both earthquakes resulting from the ISC hypocenter and GCMT focal mechanism give the smoothest synthetic data with the largest amount of data fitting. For the Kahak earthquake, the ISC hypocenter provides the best fit to the observed data with the maximum total variance reduction of 35.30 % for the spatial and 54.50 % for the spatiotemporal distribution. For the Kazerun earthquake, the best fit to the observed data with the maximum total variance reduction of 54.44 % is obtained using the ISC hypocenter. Also, the sensitivity of the slip models to some influential parameters such as rupture velocity and rise time are explored. This sensitivity test shows that increasing the rupture velocity increases the seismic moment and decreases the total variance reduction. Moreover, different values of the rise time demonstrate that the rise time growth reduces the rupture area and the seismic moment. Another result of this test is that slip distribution is heavily influenced by the number of stations and different sets of data. Finally, a comparison between the slip pattern in the fault plane and its projection on the earth’s surface illustrates that the aftershocks distribute primarily outside of the majority of the slip. Since the aftershocks are a phase of relaxing stress concentrations, they are expected to spread outside of the slip patch. For the Kahak earthquake, the distribution occurs near the western end of the slip model. There is also a predominance of aftershocks which surrounds the majority of the slip in the Kazerun model. Therefore, an analysis of the aftershock distribution and slip patterns may prove the reliability of the solutions.

Keywords

Main Subjects