Spectral decomposition and spectral balancing of seismic data

Abstract The interpretation of discrete stratigraphic features on seismic data is limited by its bandwidth and its signal-to-noise ratio. Unfortunately, well-resolved reflections from the top and base of subtle stratigraphic geologic boundaries occur only for thick features imaged by broadband data. Seismically thin stratigraphic features approaching a quarter-wavelength in thickness give rise to composite, or “tuned,” seismic reflections. Different spectral-decomposition methods provide an effective way of examining the seismic response of stratigraphic geologic features in terms of spectral components and thus help in interpretation. Phase components help with interpretation of the discontinuity features as well as stratigraphic features such as onlap, offlap, and erosional unconformities. Applications of an often overlooked attribute derived during spectral decomposition, called the voice components, can be illustrated in terms of more accurate interpretation of the subsurface features. An “amplitude-friendly” method for spectral balancing enhances the frequency content of the data and preserves the geologic tuning features and amplitudes. Spectral decomposition of seismic data that are spectrally balanced and interpreted in terms of voice components leads to more accurate definition of the features of interest.