Reduction of Bias and Uncertainty in MASW Inversion by Full Dispersion Spectrum Inversion

Summary The conventional surface waves method, which relies on fundamental-mode (FM) inversion, frequently faces challenges such as mode misidentification and the generation of multiple models that fit the experimental data equally well. Significant epistemic uncertainty, often overlooked in surface waves inversion, arises from variations in layer parameterization. This study highlights the effectiveness of the full dispersion spectrum inversion in reducing epistemic uncertainty compared to the common FM inversion. The proposed method leverages the entire dispersion spectrum by incorporating an efficient forward model. Layer parameterization is analyzed by varying the maximum resolvable depth and the layer number, with model uncertainty visualized through the spread of inverted results across different parameterizations. The inverted results show that dispersion spectrum inversion significantly reduces variability caused by parameterization differences, consistently delivering more accurate outcomes without requiring mode identification. This approach offers a robust framework for subsurface characterization using surface wave methods.