Accuracy Estimation of Earthquake Source Geometry Parameters by the Sterling Interpolation Method in Nonlinear Inversion

Abstract

Estimating the accuracy of source geometry parameters is helpful for understanding error propagation in a nonlinear inversion. To date, no studies have estimated the nonlinear inversion accuracy of source geometry parameters. Multipeak particle swarm optimization (MPSO) is used to invert source geometry parameters, providing them good reliability as is shown by Monte Carlo analysis. Also, Sterling interpolation is introduced to estimate the accuracy of nonlinear inversion of parameters in earthquakes. The square roots of variance in the source geometry parameters are obtained by the displacements and the corresponding errors. A simulation experiment illustrates that changing the number of observation points does not affect the accuracy of the seven source geometry parameters. When the number of observation points is constant, the square roots of variance in the source geometry parameters increase with an increase in noise in the observations. Finally, Sterling interpolation accuracy estimation as proposed is applied to the Lushan and L’Aquila earthquakes. We find that the random noise in the observation data affects the inversion results. The applicability and feasibility of the proposed method are verified.