Pre-Stack Enhancement: Technical Details
Residual multiples are often present in migrated gathers. They may interfere with primaries at mid to far offsets and may bias AVA attributes at near offsets. We offer a high-resolution Radon demultiple which is very effective at removing longer period multiples where there is sufficient moveout difference between the primaries and multiples. The high resolution algorithm avoids damage to primaries, allowing quite tight cuts with effective multiple suppression. For shorter period multiples we use tau-p deconvolution.
Coherent noise reduction is achieved in gathers using a linear high-resolution Radon transform. Dipping noise in the inline-crossline domain is attenuated using f-k-k filtering applied to offset cubes. Random noise is reduced by using edge-preserving spatial filters or anisotropic diffusion filters, applied to gathers or to offset cubes as necessary.
Residual velocity errors are corrected by semblance scanning the gathers. A simultaneous scan for velocity and higher-order moveout (Alkhalifah’s eta parameter) is used and these updated fields are available for QC and further analysis. Non-velocity related problems are addressed using a trim-statics method. There is very tight control on the stability of the time shifts and the algorithm design ensures that type 2 P AVA responses are not damaged by misalignment.
These techniques are applied as necessary. The order in which they are used depends on the nature and severity of the problems in the data, and workflows are designed individually for each individual dataset. Test areas are defined using the health check attribute maps to ensure that all issues are adequately addressed, and the same attribute maps after conditioning are used to QC the work.