Surface Wave Analysis in Laterally Varying Media

Abstract

Surface wave analysis applied to near surface characterisation has had a tremendous development in the last decade. From individual inversions of single dispersion curves extracted from on purpose acquired small scale datasets, the technique has evolved to the analysis of wide datasets of dispersion curves extracted from large scale seismic data acquired for body wave exploration (Socco et al., 2010). The evolution of the method poses new challenges and provides new opportunities for retrieving near surface velocity models in complex geological environments. In particular, innovative approaches for processing and inversion of surface wave data have been developed to improve the reliability and the spatial resolution of the velocity models. If data are not acquired on purpose for surface wave analysis, careful data evaluation is required before processing to assess that acquisition parameters and equipment are suitable for retrieving good quality dispersion curves. After preliminary evaluations, data should be processed to extract dispersion curves taking great care about the effects of lateral variations and the presence of higher modes and other guided waves that could be included in the inversion. Before inverting the curves, apriori information should be considered to build up consistent initial models optimising the parameterisation according to reliable investigation depth, vertical and lateral resolution. Inversion should be performed considering the set of dispersion curves as a unique dataset to provide internally consistent pseudo-2D/3D velocity models and including any available a priori information. Experimental uncertainties should be also used to account for data quality. Joint inversion with other geophysical data can significantly improve the reliability of the final models and the amount of retrieved information. Sensitivity analysis may provide an indication about reliability of estimated model parameters.