Data-driven, convolution-based surface-related multiple prediction and attenuation package
SRME includes three major components: data interpolation/regularization, multiple prediction, and subtraction.
- Data interpolation/regularization: Multidimensional data interpolation and regularization improves the surface data conditions and spatial sampling while preserving the azimuth information. This process partially overcomes the acquisition imperfections and, in turn, improves the multiple prediction quality.
prediction: To handle both wide azimuth and narrow azimuth surveys, TAME™ uses the true azimuth information for contribution trace selection to predict the multiple. True azimuth information is also used to define the aperture for each predicted output multiple trace. The program generates free-surface multiples for every source and receiver combination by a convolution process over a specified inline and crossline aperture. The inline and crossline aperture values will be tested as well as key parameters in the adaptive subtraction program.
- Subtraction: Domain decomposition adaptive subtraction, which seeks to optimize the adaptive subtraction parameters in various domains, such as frequency, dip and so on, is used to provide optimal adaptive subtraction, especially in regions of complex multiples.