This paper discusses two seismic data conditioning processes – Structural Filter (SF) and Gabor deconvolution. We found that these two techniques can improve AVO analysis in CMP and CRP gathers. Structural Filter greatly enhances the S/N of gathers by removing abnormal single high amplitude events, outliers and strong random noise. At the meantime, the relative amplitude is well preserved. A smoother AVO anomaly is observed in the gathers with enhanced S/N. The gather was measured for AVO fit using a 2-term Aki-Richards equation. The AVO gradient analysis shows an amplitude change discontinuity at far offsets. We speculated that this is caused by tuning effect. Gabor deconvolution is used to increase the resolution of the data and remove the Q effect. The gradient analysis showed a much more reasonable class 2 AVO anomaly with a fit of 0.85 for top of the layer after Gabor deconvolution.
AVO analysis contributes significantly in reservoir characterization and seismic interpretation. But AVO suffers from a wide variety of complications and problems. Castagna (1993) showed the myriad of factors affecting seismic amplitudes. According to Sherwood et al (1983) “The simple interface reflection coefficient is only a starting point toward understanding offset-dependent-reflectivity. More often we deal with reflections from layers, transition zones, or complex layered sequences.’’ A robust data processing scheme for AVO analysis needs to be carefully selected so that noise could be suppressed without biasing or corrupting the reflectivity variation with offset. Two specific AVO processing techniques that we will discuss in this paper are structural filter and Gabor deconvolution.
Figure 1: VSP data from well Wardner 175. Horizon of interest is F11 at ~1.58s. (El-Mowafy and Marfurt, 2008).