Prestack Regularization Techniques

Compact Fourier Interpolation (COMFI)
This unique technology enables a high-quality regularization solution providing a first-class input to imaging, 3D demultiple, or 4D processing. COMFI is an high-quality multidimensional interpolation algorithm, allowing spatially irregular input data to be output onto a set of user-defined spatial positions (regular or irregular). The use of local operators enables us to apply the optimum interpolation operator for each output point.

Dominant Azimuth Regularization

After reposting the streamer data to inline bin center, dominant azimuth regularization is applied in the crossline direction to bin-center the traces onto the 3D grid. This two-pass procedure combines high accuracy with efficiency.

Azimuthal Moveout (AMO)
AMO is an advanced technique for effective prestack interpolation and regularization. It can be used to correct for cable feathering in marine environments or to compensate for unevenly sampled land data. During AMO, azimuths are rotated and offsets modified, benefiting subsequent demultiple or imaging processes. In addition, AMO can be used to prepare for processes that require the input data to have zero azimuth and nominal spatial locations, e.g., for fast common azimuth depth migrations.

AMO, while originally proposed for marine streamer data, has been successfully included in the processing sequence for 3D Land or TZ surveys characterized by highly variable fold and azimuth distributions.

Flex Binning (FLEX)
The traditional approach to fold regularization, but still worthy of inclusion in processing workflows as a preconditioning step for more sophisticated interpolation techniques.

Modified Flex-Binning (Robin Hood)
This method, developed by WesternGeco, involves flex binning across offset planes within a cmp and includes a differential moveout correction to form new traces from moved offset groups. In essence, Robin Hood steals from the rich (offsets) and gives to the poor (offsets)!

Variable Offset Grouping
A process particularly suited for wide-azimuth surveys where the near and far offsets tend to be severely under sampled in the offset domain. The offset grouping need not be a fixed value and increment. The offset groups can also overlap.

Request More Information