The Sensitivity Analysis as a Non-uniqueness Indicator in the Inversion of Rayleigh-wave Dispersion Curve

Abstract

In the Rayleigh-wave dispersion curve (RDC) inversion, layer thicknesses are typically fixed during the inversion process and may be estimated using available a priori knowledge. But, the RDC inversion process frequently suffers from non-uniqueness due to the use of erroneous or conjectural layer thicknesses. In this work, sensitivity analysis of S-wave and layer thickness is used to reveal the reliability and accuracy of the RDC inversion results in the non-uniqueness situation. For this purpose, a synthetic data is inverted with three different initial models through the same inversion parameters and their sensitivity patterns are compared. The sensitivity analyses demonstrate that the sensitivity pattern of layer S-wave velocity and thickness may present useful information about the quality of estimated parameters. In this way, the possibility of getting trapped in non-uniqueness may be decreased.