Principal objectives and sub-goals
The ultimate goal of the present project is to provide original ray-based and target-oriented Prestack Depth Migration (PSDM) approaches in seismic imaging within a dynamic framework. Some keywords are modeling-driven data access and flexibility with respect to survey, imaging zone, wave type and application scale (from near surface to larger crustal structures, via O&G targets). Another important aspect is the introduction of model perturbation to take into account iterative velocity-model building (tomography) and imaging, with integrated velocity analyses using Constant-Image Point (CIP) gathers in the angle domain.
This project is funded by the Research Council of Norway as a Strategic Institute Programme (SIP, #181688/I30).
Summary
SIP will represent an important and considerable research effort with the main objective of developing a new seismic imaging system at NORSAR. Recorded or synthetic data will be input to the imaging system in order to generate target-oriented depth images of the ground structures detected by elastic waves. The imaging or migration methods to be used and further developed, are ray-tracing based, i.e., classified among Kirchhoff summation approaches or Diffraction Stack, as opposed to so-called wave-equation methods. Such Kirchhoff methods need Green’s functions (traveltimes, amplitudes, etc) to correct for wave propagation effects between all sources and receivers, and all points of the target zone to image.
Example of three local depth images (LI) superimposed on a standard Diffraction
Stack (DS) and obtained knowing only the Green’s functions at one point (red dot),
in contrary to the DS image in the background, which needs a dense grid of
Green’s functions, hence more time-consuming (Lecomte et al., 2005).
The key method to compute ray-tracing based Green’s functions is the 3D Wavefront Construction (WF3D) ray tracer developed by NORSAR since 1993 and now part of the commercial modelling packages. WF3D is indeed already used by the industry for imaging purposes. The existing WF3D tracer will therefore be adapted for an even more efficient and flexible use in imaging, and interpolation techniques will be studied to allow coarser grids of Green’s function. The on-going developments of WF3D towards complex anisotropy will also be a fundamental basis for imaging. Another class of methods to be used for Green’s funtions are Eikonal solvers, to complete WF3D with first-arrivals for specific applications.
Example illustrating the importance of a proper control of the imaging locally,
i.e., down to the image points with left) standard depth image obtained by
Diffraction Stack and right) improved depth image when applying proper
ray-based illumination-controlled imaging locally (Lecomte et al., 2005).
A second key element of ray-tracing modelling is an adequate depth model. There is indeed a need for a more flexible type of model, including a new dimension, i.e., the possibility to easily perturbe the model. This is important for integrating modelling, tomography and imaging, as all these steps are needed in a complete inversion approach. Efficient and modelling-based data access to large files on hard disks is also a necessity for target-oriented imaging. At last, modelling by demigration will be studied too because such techniques are closely related to imaging and will significantly extend the modelling tools of NORSAR.
|
