Moveout, velocity, and stacking

Moveout, velocity, and stacking

Jon Claerbout

In this chapter we handle data as though the earth had no dipping reflectors. The earth model is one of stratified layers with velocity a (generally increasing) function of depth. We consider reflections from layers, which we process by normal moveout correction (NMO). The NMO operation is an interesting example of many general principles of linear operators and numerical analysis. Finally, using NMO, we estimate the earth's velocity with depth and we stack some data, getting a picture of an earth with dipping layers. This irony, that techniques developed for a stratified earth can give reasonable images of non-stratified reflectors, is one of the ``lucky breaks'' of seismic processing. We will explore the limitations of this phenomenon in the chapter on dip-moveout.

First, a few words about informal language. The inverse to velocity arises more frequently in seismology than the velocity itself. This inverse is called the ``slowness.'' In common speech, however, the word ``velocity'' is a catch-all, so what is called a ``velocity analysis'' might actually be a plane of slowness versus time.

Moveout, velocity, and stacking