In general, the oil industry makes the time-to-depth conversion of seismic data by the image ray tracing method. This method takes time into the depth, point to point, the amplitudes of the time migrated seismic section. For each point of the migrated time section, it is necessary to trace a ray perpendicular to the surface. After this, the amplitude of the migrated point of the section takes place in depth. The seismic migration method pre- or post-stack consists of placing seismic events in the correct positions in time or depth sections. Seismic depth sections provide an image near of the subsurface, in order to facilitate the identification of possible oil accumulating geological structures. The conversion of sections from the time to the depth domain is an intermediate step in the construction of seismic images in depth. This work developed and tested a method of converting time to depth the zero-offset seismic sections. In this case, the construction of sections in depth uses normal ray tracing method. The proposed method makes use of the (slowness versus time of intersection) transformation on the zero-offset section. Each point in the domain provides initial conditions for the normal ray tracing: a start position of the initial rays and initial angles formed with the normal to the surface, i.e., the slowness initial parameter. Unlike ray image method, several rays use the same travel time and the same initial position, defining an isochronous curve. The amplitude of each point in the zero-offset section takes place to depth from the distribution of values along each isochronous curve in depth. The image ray based time-to-depth conversion has good recovery of the depths of reflectors as well as low computational cost, since it is necessary only one ray to convert each point of the section in time. However, the reflector continuity may be damaged in case of sharpened curvature. In turn, by normal ray the time-to-depth conversion correctly approximate the depth of the reflectors, since the same point in time assigns several times in depth. However, it has a higher computational cost, because it is necessary many rays to convert one point in time.
|
