Eric Verschuur

Surface-related multiple removal

Multiples are still a major problem in seismic processing, both for the marine as well as the land case.
One of the main concerns in multiple elimination is to preserve the AVO information of the primary reflections. Within the DELPHI consortium surface and internal multiple elimination algorithms have been developed, which are based on auto-convolutions of the seismic shot records. For land situations with a rather hard top layer surface-related multiples may occur, being generated between the surface and several of the subsurface reflectors. The existence of strong surface multiples is often coupled with high seismic velocities from the target up to the surface, resulting in lack of velocity discrimination.
The surface-related multiple removal can still distinguish between primaries and multiples.
Multiple removal approaches have been developed for application on poststack, CMP and shot data.
Careful preprocessing of the data is one of the pre-requisites.

Stack of data with multiples

Stack after removal of surface multiples

Removed surface multiples only

Internal multiple removal

In some situations, internal multiples can be strong as well, e.g. between top and bottom of salt or basalt layers. The method of surface multiple removal can be extended to other multiple generating boundaries.
Two options are available:
1) Interface-based: using CFP operator estimation technology to get propagation description towards the multiple generating boundary, the interface-related multiples can be predicted;
2) Layer-based: only an approximate description of propagation effects are required to remove all internal multiples that pass a certain user-defined level.
Below an example of surface and layer-based internal multiple removal on a dataset from offshore Norway is shown.
The addressed internal multiples are generated between various reflectors, but can be handled in one layer-based procedure.

Stack after removal of surface multiples

Stack after removal of surface and internal multiples

Removed internal multiples only   

CFP-based processing

Over the last years, the concept of migration in terms of double focusing has been introduced within the DELPHI consortium.
The important concept here is the CFP-gather (Common Focus Point gather): the response of a focused source (or detector) located at a subsurface gridpoint, i.e. the result of a synthesis of the surface data, with the detectors (or sources) still at the surface.
A CFP-gather can therefore be considered as half-focused data. In this domain, the principle of equal traveltime is valid: the one-way traveltimes from the focusing operator correspond with the traveltimes of the resulting reflection event in the CFP gather (see below).
As a consequence, any wrong focusing operator will produce CFP gathers in which the reflection event does not correspond in traveltime.
Applying a time correction of the CFP gather with its operator traveltimes - yielding differential time shift (DTS) panels - will show non-aligned events around zero time if the operator is wrong (see also below).
  The applications of the CFP technology are numerous:

  • Estimating focusing operator without estimating the velocity depth model
  • Velocity independent redatuming
  • Internal multiple removal
  • Imaging beyond depth migration
  • Imaging converted wave constently with PP reflections
  • Getting angle-dependent reflectivity along boundaries

Shot record in 3-reflector model and focusing operator for 2nd boundary

CFP gather for point at second boundary and resulting DTS panel

Dr. ir. D. J. Verschuur