oa Converting CRS attributes to interval velocities: An approach to build an initial velocity model for pre-stack depth migration

Pre-stack depth migration is proven to enhance the seismic image but only when accurate velocities are used. Conventional methods for building an initial velocity model fail in the presence of structural deformations and severe lateral velocity variations. Therefore, for depth imaging of complex structures, a different approach is required. The velocity attribute derived from the common reflection surface (CRS) technology offers an alternative method to build an initial velocity model for such a geological setting. In this presentation I propose a methodology that exploits the CRS attribute values to estimate interval velocities. The CRS method is based on optical concepts and indirectly provides root-mean square (RMS) type velocities. It is data-driven and honors horizontal and vertical velocity changes. It also provides continuously sampled results with better data statistics, but the conversion to interval velocities is still not straightforward. One of the methods to handle the high variability of the CRS attribute is to use geological<br>constraints for editing and smoothing. The editing of extreme values is done according to a general velocity trend, while smoothing utilizes a structural model provided by the interpreter. Therefore, it is possible to convert CRS attributes to plausible interval velocities. The proposed methodology was successfully applied to obtain initial macro-model velocity estimates for several onshore 2-D lines acquired in complex salt tectonic geology, near the Red Sea of Saudi Arabia. As a result, the required number of pre-stack depth migration iterations was reduced, and the final seismic image was enhanced.