Imaging the Subsurface with 2-D and 3-D Seismic Data

Abstract

In regions of complex geology, the corresponding seismic time sections show diffraction events, distorted positions of reflectors, smeared lateral discontinuities and uncertain amplitudes of seismic reflections with the consequence that the interpreter cannot automatically reconstruct and identify the cross-section through the earth from the seismic time section. The purpose of migration (imaging of seismic data) is to reconstruct structural depth sections from seismic time sections and to produce a geological image of the subsurface. ‘Poststack migration’ is an imaging process that uses stacked averaged seismic data. Stacking reduces the amount of data drastically, and thus subsequent migration of stacked data is less expensive than prestack migration. On the other hand, severe problems occur where the standard stacking procedure can no longer be applied, e.g. when the reflection surfaces are strongly curved or there are strong variations of velocities. The ‘prestack seismic migration process’ is a reconstruction of subsurface structures directly from measurements at the surface of the earth. The advantages of this process particularly appear in faulted steep dip regions with extreme flexures. Furthermore, this process allows an iterative estimation of the migration velocity field (macro velocity field), which is needed as an important parameter field for the migration process itself. ‘Depth migration’ is a transformation, mapping, and imaging of measured seismic time data into a migrated depth section. It yields a reliable image of the subsurface, provided that the migration velocity field (macro velocity field) is correct. ‘Time migration’, in principle, represents a special type of depth migration, where the depth scale τ is scaled by a pseudo-time scale using a replacement velocity. Most time migration schemes do not take into account the refraction of waves at the boundaries, hence, these schemes fail to image data from laterally inhomogeneous media. On the other hand, time migration schemes show low sensitivity to migration velocity errors and can be used as an intermediate step in seismic data processing for imaging seismic data. The conventional seismic 2-D technique provides an erroneous image of the subsurface, if the subsurface varies perpendicularly to the 2-D seismic recording direction. In this case, a 3-D technique has to be applied. The 3-D migration process plays a fundamental role in the 3-D processing sequence. It provides a reliable image of the earth and avoids possible misinterpretation which may occur when 3-D structures are explored with 2-D methods. 3-D poststack time migration is often used in practical work because this first result is needed for first interpretation and to check the effect of 3-D processing and 3-D imaging. Nevertheless, this type of imaging fails to give a reliable image of the subsurface in the case of complex geological areas, e.g. complex overburden structures. 3-D poststack depth migration needs the same computing time as 3-D prestack time migration, but yields a reliable depth image of the subsurface provided that the conditions for 3-D stacking are given and a reliable 3-D macro velocity field is available. 3-D prestack depth migration has long been recognized as the appropriate imaging procedure for complex structures and complex velocities. It requires that the velocity field is estimated using the migration procedure itself in an iterative manner, however, therefore this migration process consumes an enormous amount of computing time. In this paper, we focus on the description of a 3-D imaging process which uses a special type of 3-D prestack depth migration only to estimate the 3-D macro velocity field and then applies a 3—D poststack depth migration scheme. This scheme is applied to a 3-D data set acquired at the Costa Rica convergent margin.