Combination of Artificial Neural Network and Genetic Algorithm To Inverse Source Parameters of Sefid-Sang Earthquake Using Insar Technique and Analytical Model Conjunction

In this study, an inversion method is conducted to determine the focal mechanism of sefid-sang fault by comparing interferometric synthetic aperture radar (insar) technique and dislocation model of earthquake deformation. to do so, the sentinel-1a acquisitions covering the fault and its surrounding area are processed to derive the map of line of sight (los) displacement over the study area. then, using the ascending and descending tracks of the satellite, the three-dimensional displacement field is recovered over the region. the maximum horizontal and vertical displacements are about 12 cm and 5 cm respectively. the resulting displacement field is compared with okada half-space dislocation model of earthquake to determine the focal mechanism and fault parameters by a nonlinear inversion method, which is composed of artificial neural network (ann) and genetic algorithm (ga). the coulomb stress and strain changes, which are important factors for prediction of aftershock event, are also determined. the numerical achievements show a slip of 4.5 mm, a depth of 8 km, dip angle of 55 deg and width of 10 km for this fault.

Combination of Artificial Neural Network and Genetic Algorithm to Inverse Source Parameters of SefidSang Earthquake Using InSAR Technique and Analytical Model Conjunction

In this study, an inversion method is conducted to determine the focal mechanism of SefidSang fault by comparing interferometric synthetic aperture radar (InSAR) technique and dislocation model of earthquake deformation. To do so, the Sentinel1A acquisitions covering the fault and its surrounding area are processed to derive the map of line of sight (LOS) displacement over the study area. Then, using the ascending and descending tracks of the satellite, the threedimensional displacement field is recovered over the region. The maximum horizontal and vertical displacements are about 12 cm and 5 cm respectively. The resulting displacement field is compared with Okada halfspace dislocation model of earthquake to determine the focal mechanism and fault parameters by a nonlinear inversion method, which is composed of artificial neural network (ANN) and genetic algorithm (GA). The coulomb stress and strain changes, which are important factors for prediction of aftershock event, are also determined. The numerical achievements show a slip of 4.5 mm, a depth of 8 km, dip angle of 55 deg and width of 10 km for this fault.