On the influence of different misfit functions for attenuation estimation in viscoelastic full-waveform inversion: synthetic study

Abstract

SUMMARY Estimating subsurface attenuation distribution is essential to compensate the amplitude and phase distortions in seismic imaging and characterize attenuative reservoirs. Full-waveform inversion (FWI) methods represent promising techniques to invert for both velocity and attenuation models with arbitrary spatial distributions. However, simultaneously determining velocity and attenuation properties introduces the problem of interparameter trade-off in viscoelastic FWI. Ignoring attenuation effects can result in inaccurate velocity estimations. Velocity errors may produce significant parameter crosstalk artefacts in the inverted attenuation models. An appropriate misfit function measuring specific seismic attribute is essential to capture the influence of attenuation on the seismic data and thus is expected to reduce the influences of velocity errors for attenuation estimation. In this study, we evaluate the performances of different misfit functions for attenuation estimation in viscoelastic FWI accompanied with a two-stage sequential inversion strategy. Synthetic examples with different acquisition surveys are given to show that in the presence of strong velocity errors, the amplitude-based misfit functions, including envelope-difference, root-mean-square amplitude-ratio and spectral amplitude-ratio, can invert for the attenuation models more reliably, compared to the waveform-difference and instantaneous phase misfit functions. With the two-stage inversion approach, more reliable velocity and attenuation models can be obtained using viscoelastic FWI.