Multiple ground-coverage seismic techniques are based on a few key assumptions. Among the more important are the assumptions that subsurface reflectors locally are continuous and linear, and that the primary reflections between particular source-receiver pairs travel along unique paths. However, overenthusiasm concerning recent advances in the use of "velocity spectra" models has led to some violations of these important assumptions. The most common types of violations involve use of these models in (1) determining primary velocity, (2) computing interval velocities and dips, and (3) migrating depth sections. Although all three processes are industry-wide objectives which commonly are obtainable by other methods, their attainment through the use of the "velocity spectra" odels is beyond the limits of current theory. Therefore, those who use "velocity spectra" methods to predict subsurface conditions can be misled by the errors which result.

The theory and limitations of the seismic methods currently employed can be clarified by examining linear

End_Page 557------------------------------

and nonlinear computer-simulation models. Ideal time sections for some of these models illustrate phenomena such as dip reversal, loss of domal character, incomplete unconformity contacts, and creation of faults. Modeling also is becoming increasingly important in reflection-seismic processing and interpretation. However, true progress with models will not take place until we appreciate and understand the limitations of the method, the primary assumptions that are essential, and the consequences of violating these primary assumptions.

End_of_Article - Last_Page 558------------