On Frequency and Resolution of Viscoelastic Least-Squares Reverse Time Migration

With the development of seismic imaging, least-squares reverse time migration (LSRTM) has been developed as one of the most powerful imaging method used for iteratively inverting the high-wavenumber components of subsurface lithological parameters and images. Taking attenuation into account, viscoelastic LSRTM are anticipated to effectively compensate the amplitude loss and phase distortion. In this abstract, viscoelastic forward modeling was conducted with different wavelets to produce observed data with a variety of frequency range. Viscoelastic LSRTM was implemented to the observed data sets, and the inverted images was analyzed to establish the relationship between the inversion resolution and data frequency in an intuitive way. The analysis shows that the resolution limits of viscoelastic LSRTM is strictly confined to the frequency components in observed data. Low frequencies in the observed data is much important for that the inversion with low frequencies can effectively fill the wavenumber gap between the background model and the perturbations. Viscoelastic LSRTM with abundant low frequency data converge faster than the ones lack of low frequencies.