A Study on the Consequence of Surface Wave Dispersion Uncertainty on VS,30 and Soil Amplification Considering Different Inversion Approaches

The paper studies the consequence of surface wave dispersion uncertainty on VS,30 and soil amplification. The current practice of surface wave methods is to carry out inversion targeting to generate a single best fitting profile with a minimum misfit. However, if the target dispersion curve is associated with some amount of data uncertainty, targeting to achieve the lowest minimum misfit may not serve the purpose. The generation of profiles covering the entire uncertainty bound should also be accounted for at the same time. Here an attempt has been made how this problem can be addressed using different inversion approaches, one is better in optimization and other represents the uncertainty bound well. To carry out other subsequent analyses, few profiles have been selected from the entire assigned uncertainty bound as it is not practical to select such a large number of initial realizations. Then VS,30 values have been estimated from those selected profiles using the concept of travel time average. Linear elastic ground response analysis has been performed and the parameters (peak amplifications and frequencies) are mapped with respective dispersion misfit values. The study combines different inversion approaches to generate optimized and uncertainty consistent profiles. It is also observed that these better optimized and uncertainty consistent profiles can provide information covering a large range of dispersion misfit values and can also carry valuable information on the convergence of the non-unique nature of inversion.