Least-squares reverse-time migration with a wavefield-separation imaging condition

Although least-squares reverse-time migration (LSRTM) can improve the resolution of the conventional reverse-time migration (RTM) image, directly imaging steeply-dipping fault zones remains a challenge when the acquisition aperture is limited. We develop a least-squares reverse-time migration method with a wavefield-separation imaging condition to enhance imaging steeply-dipping fault zones. Unlike the conventional imaging condition that includes the total source and receiver wavefields, our LSRTM with the wavefield-separation imaging condition cross-correlates the wavefields propagating in opposite vertical or horizontal directions. The conventional imaging condition used in LSTRM uses a fixed source wavefield calculated using the initial model for all subsequent iterations, which may not be able to clearly image fault zones. Our new LSRTM accounts for the updated source wavefield in each iteration. Synthetic examples demonstrate that our LSRTM significantly improves the conventional LSRTM image for imaging steeply-dipping fault zones.