Taner, M. Turhan1, Sven Treitel2
(1) Rock Solid Images, Houston, TX
(2) TriDekon, Inc, Tulsa, OK
Estimation of seismic attenuation is as important as the estimation of interval velocities. Attenuation provides an additional perspective of the lithology and reservoir characteristics. High frequency losses due to absorption reduce the bandwidth of seismic, and consequently the resolution of the seismic images. In one case, we need to determine attenuation, while in the other we have to compensate for it. Due to the low frequency content of seismic waves, the attenuation effects are usually small. They can be observed with accuracy over only large depth intervals. In this paper we show a computational procedure that allows us to estimate the seismic Quality Factor (Q) and design inverse filters to compensate for attenuation. Spectral ratios from such inverse filters are then used to determine Q.
We have derived a robust Q estimation method that is based on determination of the unit step-ahead prediction error operator for an arbitrary seismic wavelet, from which its minimum phase equivalent wavelet can be obtained by synthetic polynomial division. By carrying out a series of such calculations along a seismic recording, we have been able to develop a new version of the familiar spectral ratio method. Numerical experiments carried out to date suggest that this approach has a number of desirable features: it is fast, computationally stable, and less prone to the influence of zeros on or near the unit circle in the z-plane because our method makes use of minimum-phase wavelets alone, which by definition cannot have zeros on the unit circle.
AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.