Stacked

Abstract

Analysis of prestack P‐wave seismic data yields information about both the P‐ and S wave properties of the earth. An anticipated advantage of having two measurements (P and S) is that they can be combined into a new measurement that is less sensitive to lithology variations and more sensitive to fluid effects. The amplitude‐variation‐with‐offset (AVO) gradient is one such measure that is often used qualitatively as a fluid indicator. The gradient always becomes softer (more negative) when hydrocarbon replaces brine in the pore spaces but the overall AVO response is dominated by the lithology. Fluid effects are expressed primarily by the normal‐incidence P‐wave response and only secondarily by the offset dependence. The gradient often does not function as an effective fluid indicator. This is partially due to the fact that the gradient is roughly twice as sensitive to S‐ than to P‐wave properties. More importantly, effective random noise in the CMP gathers introduces a strong correlation between the AVO intercept and gradient and, hence, between the measured P‐ and S‐wave properties. This correlation in the AVO attributes corresponds to a significant error in the estimation of the S‐wave properties and can dominate the measurements from many of the popular AVO techniques. A simple method to minimize the effect of this noise‐induced correlation is to stack the data. The stack corresponds to a coordinate rotation in elastic space with the stack amplitudes measured along one of the new axes and the other (unmeasured) axis naturally tending to line up with the noise and thus suppressing it. Fluid effects cause the data to move roughly perpendicular to this noise trend. The stack axis is then in the direction of the fluid effect. The stack thus combines both the P‐ and S‐wave (normal and oblique incidence) information into a single measurement which can be made to optimally suppress background noise and highlight fluid effects. A major consequence of this interpretation is the simplicity of both prospect identification and quantitative amplitude analysis.