Wave Imaging Technology Inc.

WIT has the capability to efficiently and effectively simulate 3D seismic data in the acoustic limit, with high quality pseudospectral and finite-difference modeling algorithms. These algorithms lie at the core of our WIT Sim service. Whether you want to model wide-azimuth marine data over your subsalt discovery, use areal shots over your land prospect to understand fracture analysis, or QC possible problems with an upcoming offset VSP, WIT can provide you with a high-quality solution.

Theoretically, finite difference approximation of the divergence should yield a faster algorithm than the pseudospectral method of computing the divergence in the Fourier domain. However, pseudospectral methods need only sample the wavefield at 2 grid points per wavelength, versus 3 points per wavelength with finite difference codes. Thus, for large models, finite difference algorithms may require domain decomposition over multiple CPUs, and efficiency never scales linearly with the number of CPUs.

Algorithm specifications and highlights:

Can handle isotropic, VTI, TTI, and HTI media
Can model marine, land, VSP, and OBC data
Shots and receivers can exist on any topographic datum
Simulates variable density as well as variable velocity
Density field can be computed from velocity using Gardner's relation for sediments
Predetermined table of stability information to fix optimum time step
Simulates up to order 50 in time step - 4th order optimal in most cases
Shots and receivers may lie between grid points



	Whitepaper: 3D Acoustic Wavefield Simulation WIT has the capability to efficiently and effectively simulate 3D seismic data in the acoustic limit, with high quality pseudospectral and finite-difference modeling algorithms. These algorithms lie at the core of our WIT Sim service. Whether you want to model wide-azimuth marine data over your subsalt discovery, use areal shots over your land prospect to understand fracture analysis, or QC possible problems with an upcoming offset VSP, WIT can provide you with a high-quality solution. Theoretically, finite difference approximation of the divergence should yield a faster algorithm than the pseudospectral method of computing the divergence in the Fourier domain. However, pseudospectral methods need only sample the wavefield at 2 grid points per wavelength, versus 3 points per wavelength with finite difference codes. Thus, for large models, finite difference algorithms may require domain decomposition over multiple CPUs, and efficiency never scales linearly with the number of CPUs. Algorithm specifications and highlights: Can handle isotropic, VTI, TTI, and HTI media Can model marine, land, VSP, and OBC data Shots and receivers can exist on any topographic datum Simulates variable density as well as variable velocity Density field can be computed from velocity using Gardner's relation for sediments Predetermined table of stability information to fix optimum time step Simulates up to order 50 in time step - 4th order optimal in most cases Shots and receivers may lie between grid points