The relation of seismic data to stratigraphy can be understood conceptually starting from a rather naive point of view. Taking a single seismic trace, we may assume that each reflection event consists of a simple symmetric wavelet and denotes a change in lithology corresponding to a change in acoustic impedance. The polarity and size of the reflection events allows the development of an underlying reflectivity series and acoustic impedance log. These types of seismic derived Information when interpreted from trace-to-trace and in light of geological principles and available subsurface Information permit accurate correlation of seismic data with the known subsurface Information where these coincide and extrapolations elsewhere.
A variety of difficulties were cited which represented departures of real seismic data from the ideal data of the naive approach. Seismic events which did not correspond to lithology were considered such as multiples and diffractions. The interpretational complexities introduced by geometric effects were illustrated, and actual propagating waveforms which differed significantly from the ideal symmetric waveforms were shown. For each such problem some technique of seismic data processing was offered as a mechanism by which the real data could be transformed or improved to have characteristics approaching the ideal.
Most of the processing procedures for such goals and other techniques in general used to acquire and process good quality seismic data are well known and so were not discussed. The technology and methods of wavelet transformation were of more recent development however, and an exposition in some depth was provided. Many illustrations