Rayleigh wave modeling: A study of dispersion curve sensitivity and methodology for calculating an initial model to be included in an inversion algorithm

• Influence of the Poisson ratio in the dispersion curve sensitivity study.
• In some cases, the dispersion curves are significantly affected by known parameters.
• In one case, the dispersion curves aren't affected by unknown parameter.
• Initial model implemented reduces problems found in the dispersion curve sensitivity.
• Importance of this work and of the priori information for successful data inversion.

This paper presents a study on Rayleigh wave modeling. After model implementation using Matlab software, unpublished studies were conducted of dispersion curve sensitivity to percentage changes in parameter values, including S- and P-wave velocities, substrate density, and layer thickness. The study of the sensitivity of dispersion curves demonstrated that parameters such as S-wave velocity and layer thickness cannot be ignored as inversion parameters, while P-wave velocity and density can be considered as known parameters since their influence is minimal. However, the results showed limitations that should be considered and overcome when choosing the known and unknown parameters through determining a good initial model or/and by gathering a priori information. A methodology considering the sensitivity study of dispersion curves was developed and evaluated to generate initial values (initial model) to be included in the local search inversion algorithm, clearly establishing initial favorable conditions for data inversion.