Investigation of causative fault of Bam (2003) earthquake using joint inversion of InSAR and strong ground motion data

The study of recognition of causative faults of earthquakes and prediction of the physics of these earthquakes is a favorite topic of earth science researchers. The parameters of an earthquake causative fault could demonstrate attributes of tectonic plates and provide critical information about the earthquake hazard assessment.

Methodology and Approaches:

 In the first step, we need to determine the fault displacement obtained from InSAR and strong ground motion data. This displacement could be measured using the observations of radar satellites via the synthetic aaperture radar interferometry method. This method can be used to map the deformations created on the ground by earthquakes (Okada, 1985). Using SARS cape software, the displacement is extracted from interferograms that are obtained from InSAR data. To obtain the displacement from strong ground motion data, the BASCO code is used to extract static displacement from acceleration data. To find the parameters of causative fault as analytical models, the joint inversion processing is aimed such that to minimize the difference between the observed and predicted data in a least-squares sense. The non-linear inversion is based on the Levenberg-Marquardt (Marquardt, 1963) minimization algorithm. It has multiple restarts to reasonably guarantee the convergence of the cost Function to the global minimum. Finally, the synthetic accelerations are created with finite fault model to demonstrate the best fit between the observed and synthetic accelerograms at stations.

The results of this study show that two faults are responsible for breakdowns in Bam city. One of these faults has transferred stress to another one that reaches to surface. The second fault is a right-lateral strike-slip fault with strike 358°, dip 84° and rake 178° to the east that has maximum moment release with 2.49 m slip. The first fault is laid beneath the second fault with strike 172° and dip 45° to west. The stress transfer pattern of the first fault on the second one is comparable with slip distribution on the second fault. Using forward modeling of strong ground motion data and fixing parameters obtained from joint inversion of InSAR and strong ground motion data, the estimated epicenter of the earthquake is determined to be 58.27° and 29.05°