Base and sub salt seismic imaging is still an unresolved issue. To solve this problem both improved processing algorithms and acquisition geometries have been heavily researched the latest years. Reverse-time migration, full waveform inversion and wide azimuth acquisition among others are techniques which may help the salt imaging. However, there is less effort trying to explain what makes sub salt so difficult to image. To do so we should go back to the geology. It is a common assumption that salt is homogeneous with constant velocity to simplify the processing of seismic data. However, there may be several complex structures and/or rapid velocity changes within the salt body, which can affect the wavefield propagation in a such way that we get severe imaging problems.
Imaging results of a 2D line in the North Cape Basin show no sign of the base salt reflection or any other sub salt reflectors. In this work we focus on the base salt imaging problem by trying to recreate the effects in the seismic data and the images. We are testing three different models which severely distort the data and the resulting images: the first with diffractors at the top salt, the second with velocity perturbations in the salt body, and the third with a combination of both effects. Comparing shot gathers and migrated images from synthetic and real data show that velocity perturbations in the salt distort the wave propagation to such extent that base salt reflector vanishes.
Seismic imaging of complex salt structures and sub salt sediments still remains a challenge. There is a strong ongoing effort in the geophysical community to improve the way we both acquire and process the seismic data from these areas. Problems with salt imaging are often related to either getting a clear image of the base salt, or of the salt flanks. This may be due to non-suitable and dip-limited migration algorithms. Farmer et al. (2006) showed the usefulness of applying twoway wave equation migration algorithms, e.g. reverse-time migration, which handles both turning waves and multi-arrivals. Better acquisition geometries may also improve the salt and sub salt imaging. Areas which are poorly illuminated with conventional geometries are better illuminated with wide- or fullazimuth acquisitions (Regone, 2006). High-quality imaging of seismic data from complex geology requires an accurate model of the velocity field in the subsurface. In real life this may not be achievable. A depth velocity model is usually obtained using a depth tomography method on seismic data (see e.g. Cutler et al., 1984). However, building models of salt bodies (or other intrusions) is often related to finding the correct shape of the salt body after sufficiently accurate background velocities are found. A trial-and-error approach is usually used to delineate the salt body. Another method which has been heavily researched is full waveform inversion, where e.g. Pratt and Stork (2006) have shown promising results in recovering the velocity model for a synthetic salt example. This technique uses the full wavefield to iteratively update the velocity model.
