Laterally constrained inversion of ground roll from seismic reflection records

Abstract Seismic reflection data contain surface waves that can be processed and interpreted to supply shear-wave velocity models along seismic reflection lines. The coverage of seismic reflection data allows the use of automated multifold processing to extract high-quality dispersion curves and experimental uncertainties in a moving spatial window. The dispersion curves are then inverted using a deterministic, laterally constrained inversion to obtain a pseudo-2D model of the shear-wave velocity. A Monte Carlo global search inversion algorithm optimizes the parameterization. When the strategy is used with synthetic and field data, consistent final models with smooth lateral variations are successfully retrieved. This method constitutes an improvement over the individual inversion of single dispersion curves.