By applying dynamic corrections a seismic trace recorded at a distance x from the energy source should be varied in such a way as to obtain a trace which would be recorded at zero‐distance, i.e. at the source itself. Only such a zero‐offset‐trace contains the correct sequence of reflection coefficients (reflectivity function), whilst all other traces contain a distorted reflectivity function. In the simplest case, the reflectivity function is compressed over a shorter time whereas in more complicated cases a partial inversion of the reflectivity function results. This happens when some of the reflection hyperbolae intersect one another. The reconstruction of the true zero‐offset reflectivity function by the application of dynamic corrections can only be an approximative process. In the first case mentioned we must expect a decrease in accuracy of the corrected trace in comparison with a zero‐offset‐trace. In the second case, where intersections of the hyperbolae occur, accurate reconstruction is clearly impossible. The problems are discussed with the help of theoretical and practical examples.

No abstract

The purpose of the seismic processing step "migration" generally is to present a reliable• section of the subsurface with respect to the correct spatial location of reflecting elements. The procedure is usually performed in the time or frequency domain. The transformation to the depth domain requires the knowledge of the underlying velocity model. A simple depth conversion of the time scale is a very limited procedure and fails completely in the presence of dipping overburden layers. Substantiallatera1 velocity variation already falsifies the result of time migration as refraction of rays is normally not considered in this process. The error depends on the amount of refraction and the depth interval between the refracting and the reflecting interface. If steep and/or conflicting dips are involved in the data a special dip moveout (DMO) processing is required to improve the stacked data for migration. Wave theoretical depth migration takes the effect of the refraction of rays into• account by incorporating the so-called thin-lens term in the migration algorithm. This technique solves both the imaging and lateral positioning problems. For evaluation of a proper depth dependent velocity field an interactive procedure is suggested by applying a horizon migration based on a ray tracing method. The resulting velocity distribution is then used for the wave equation migration of seismic data leading to more reliable depth sections. The effectiveness of the method is illustrated by a sequence of field examples.