Tying well tops to seismic reflections requires detailed discussion of time-to-depth and depth-to-time conversion, as well as migration to true depth. Historically, tying wells to a seismic migration was called depthing and usually converted only a few horizons to depth. This depthing process was almost solely focused on ultimately providing an accurate conversion of an interpreted migrated time map to depth in such a manner that the horizon depth in all wells within the mapped area were matched as precisely as possible. This process was not concerned with producing depth volumes containing all horizons of interest; it was only concerned with matching well tops. Accomplishing this feat required that all well tops of interest be tied to the corresponding seismic time image with precise accuracy. When necessary, well velocities were modified to fit the observed discrepancies, and, after much trial and error, a suitable velocity v0(x;y) map for depth conversion of the given horizon was produced. The underlying assumption in all of this was that the time migrated volume was as accurate as necessary, and well discrepancies were just a function of measurement error. It is now known that the truth is not so simple. The reason we had to tie the seismic data to the well was that the migration was performed with no consideration for anisotropic wave propagation.
The single most important parameter associated with anisotropic migration is the 3D Earth model. Note that it is not just the simple velocity model normally used in prestack time (RMS) or prestack depth (INTERVAL) migrations. This model has at least three, and up to five, interrelated parameters.
The best known elements of this five-member set are the sound speed, determined through iterative migration velocity analysis, and the well-based or true depth velocity field. Frequently, these two fields are considered to be independent and totally unrelated, but the theoretical facts are completely out of phase with this concept. In fact, these two fields are the most important aspects of what is required to produce migrated images that exactly tie the wells.
While the two velocity fields are necessary, they are not sufficient to complete the process. In addition, we must also determine how these two fields are related to accurately image the recorded data at the precise subsurface location from which they were reflected. In the sections on velocity analysis and velocity model building, we develop and discuss the data, workflow, tools, and concepts necessary to construct the entire Earth model.
There is no doubt that there are still many geophysicists who believe that true depth imaging is not possible and will never be possible. What we argue is that all of these approaches have their place, but in the final analysis, the optimum approach must incorporate the full anisotropic model to achieve true depth conversion at all dips. Integration of all available data is key and must be performed accurately for this to produce high quality results.
