Work to be done

In order to prove beyond reasonable doubt that an inversion is feasible, an actual extraction of the FEAVO anomalies in angle gathers should be performed. This should not pose any problems, amounting in the end to image processing - discriminating the anomalies based on their morphology. Since the third prerequisite for an inversion is an invertible operator, we should also show, at least on a particular case, that the linearized downward continuation operator preserves the FEAVO anomalies. Thus, after proving that an inversion for the velocity model that produces the FEAVO anomalies is possible, the work of setting up such an inversion first for 2D, then for 3D, remains to be done. The operators must be constructed and a reliable discrimination method in angle gathers set up.

We will approach the solution of the problem gradually, incorporating elements of complexity one by one. We first have to develop a process that will successfully separate the FEAVO anomalies in the angle domain, making them ready for input into an inversion scheme for obtaining $\Delta s$. We will then implement a 2D synthetic case, then a 2D real. We will develop the theory for the 3D case, build a 3D synthetic, implement a 3D synthetic case, then a 3D real one, and finally lay out the results in a publishable form.