Smooth velocity models in reflection tomography for imaging complex geological structures

Abstract

Seismic imaging of geological structures with severe lateral velocity variations requires pre-stack depth migration of the seismic data. Such processing requires itself an accurate determination of the distribution of the propagation velocities. Reflection tomography turns out to be quite attractive for this purpose. Furthermore, migration velocity analysis can complete this technique whenever complexity of wave propagation makes the picking of reflection traveltimes very cumbersome, if not impossible. Two different subsurface representations can be used to perform these methods: the blocky model representation and the smooth model representation. In reflection tomography, using blocky models with finite velocity jumps can create shadow zones and the possible non-definition of the forward problem. Smooth models, on the other hand, are created such that they do not have such shadow zones but require specific techniques to integrate a priori geological information. Also, use of blocky models for migration brings, in general, artificial dicontinuities to migrated seismic events, thus making almost impossible the interpretation of these events which is the basis of migration velocity analysis. Should we use smooth models, such an interpretation becomes possible. Thus, in spite of some inherent limitations, smooth models are well adapted to run reflection tomography coupled with migration velocity analysis for the imaging of complex geological structures.