Paper Summary
Significant advancements have recently been achieved in the use of full waveform inversion (FWI) methods for the derivation of high-resolution model parameters. In comparison to implementations using the acoustic approximation, it has been shown that accounting for elastic effects through the use of elastic equations significantly enhances the quality of inversion results. Additional improvements can be achieved by employing variable density parametrizations; however, the low frequency component of the density model is notoriously challenging to constrain properly. An alternative approach is aiming for the reconstruction of high frequency components of the density model, which can be achieved by reparametrizing the elastic equations in term of impedance or reflectivity. We demonstrate the benefits of this new parameterization by comparing synthetic data obtained from elastic impedance inversion to field data from an OBN survey.