Multi-Point-Source Describing Large Earthquake Based on Rupture Process Inversion and Image Segmentation

Abstract

The rupture process of large earthquakes is generally complex and contains multiple sub-faults planes with different focal mechanisms. The focal mechanisms inversion of these sub-faults by applying the Multi-Point-Source Faulting Representations (MPSFR) are essential for seismic stress analysis and earthquake disaster assessment. The MPSFR method is time-consuming and often with unstable results. In this study, we develop an Image Segmentation and Iterative Inversion (ISII) approach to calculate the MPSFR for large earthquakes by inverting near-field strong motion data. This new approach analyzes the rupture image of the earthquake and divides the entire rupture surface into several sub-rupture segments as a point source in the MPSFR. We approach the ISII model to the 2010 El-Mayor Cucapah (EMC) earthquake and the 2016 Kaikoura earthquake, respectively. In the EMC earthquake, the overall misfit was reduced from 0.58 (earthquake rupture model with the same focal mechanism) to 0.47 (IISI model with four different focal mechanisms). In the Kaikoura earthquake, the overall misfit was reduced from 0.67 to 0.55. The rupture process inverted by the ISII model is consistent with the joint multi-method inversion and the operation process is high efficiency. The test results indicate the ISII model can accurately and quickly invert the complex earthquakes rupture process and provide valuable information for earthquake disaster assessment.