The Application of Optimal Wavelets in Seismic Inversion

The current emphasis in seismic stratigraphy has This process of wavelet extraction, followed by the focussed on velocity inversion and the generation of derivation of the reflectivity sequence, allows the the pseudo-acoustic log from the seismic trace. ideal pre-conditioning for seismic velocity inverThere are several published methods on wavelet sion. extraction which invariably make assumptions about the wavelet, in terms of the phase spectrum, The significance of this approach is demonstrated as well as assumotions about the reflectivity. in a series of synthetic and real data examoles. The The SONISEIS method will provide the basic statistically computed wavelets allow for synthetic wavelet using a statistical estimation, indepenseismograms which consistently outmatch those dently of any assumptions about the wavelet. The with Ricker or best-estimate zero or minimumderived wavelet is re&ced in the seismic trace phase wavelets, with a spike of approphate amplitude and polarity. Although the computation of velocities from reflection seismic data has been a standard practice for many years (Green 1938, Dix 19551, inversion of the seismic trace, or the reduction of a seismic trace to a sampled log of interval velocities in the transmitting medium, was first introduced in 1970 at a meeting of the European Association of Exploration Geophysicists in Edinburgh (D&s et al 1970). This approach was subsequently discussed by Lavergne (1974) and Lindseth and Street (1974) and several companies have introduced their version of this technique over the last year. Derivation of interval velocities from seismic reflection amplitudes can be viewed as the next evolutionary stage to ‘bright spot’ analysis. The development of this processing technology has historically been hindered by several factors; lack of true amplitude (for true reflectivity), severe filtering in the field resulting in limited bandwidth and possibly a lack of appreciation of the seismic implications in exploring for stratigraphic traps. In exploration for structural traps, the relative amplitude of reflections and the interval velocities of various layers are of little significance. The large scale stratigraphic variations can often he deduced from velocities computed for normal moveout corrections. However, the exploration for subtle stratigraphic traps focussed interest on bright spots in the early seventies and drew attention to the relationship between reflection amplitude and the velocity of the material across the interface. It was quickly real&d that the resolution of normal seismic reflection events spanning 30.60 msecs. on a. trace was inadequate to define layers even up to 100 feet thick. This desire for greater resolution spurred heightened interest in detailed inversion techniques. The integration of a normally processed seismic trace, even after phase compensation, is not enough Presented at the 47th Annual Meeting, September 14th, 1977 in Calgary. Publication author&d by the Editor, Geophysics. ’ Commonwealth Geophysical Development Company Ltd., Calgary, Alberta “PanCanadian Petroleum Limited, Calgary, Alberta