Tomographic Pre-Stack Depth Migration (PSDM) – Layered Earth Velocity Model
The 2D, 3D-velocity model consists of a parameterized, interval velocity within a structural model derived from a structural interpretation of the data. The parameterized velocity within each layer of the model is most commonly described as either constant interval velocity or a Vo+kz function at each x-y location. This parameterized velocity field for each layer can vary in the horizontal direction.
To update the velocity field, pre-stack depth gathers are generated for a set of target lines over the project area. Residual depth delays at horizon boundaries for all layers are extracted from the pre-stack gathers along each target line. These delays drive the analysis. Both Deregowski and tomographic analysis methods are available for updating the velocity fields.
With good data and a good current velocity/structural mode, the tomographic method can be used to update the entire model at once. Iteration of this analysis continues until the pre-stack depth gathers are flat for the horizon events and the model boundaries fit the seismic events. For very complex structures and/or poor data, it is often necessary to develop the model from the top down (layer stripping) through the difficult areas, after which one may revert back to the whole model tomography approach.
The Deregowski adjustment may also have to be used in lieu of tomography in these cases. Tomography honors the lateral changes in velocity, while the Deregowski method does not honor lateral changes at each xy location. Lateral changes are accommodated for the Deregowski method through interpolation of the velocity field throughout the volume using all of the delay analysis results. Again, iteration of this analysis continues until the pre-stack depth gathers are flat for the horizon events and the model boundaries fit the seismic events.
Diagrams show the input to Tomography and how the solution converges after Tomography.