Abstract
Modeling seismic propagation is critically important to our work; unfortunately, we often must trade simulation accuracy for reduced computational expense. We present a new seismic-modeling method that is as simple and computationally efficient as Snell’s law ray tracing but provides propagation paths and arrival times more consistent with finite-bandwidth data. We refer to this modeling method as wave tracing and apply it to nonlinear traveltime tomography and depth imaging. By replacing Snell’s law ray tracing with wave tracing, we get better ray coverage, more robust and faster ray bending (fewer iterations), and a much more robust and faster algorithm for nonlinear tomography (fewer iterations, too). A very significant benefit is increased stability and robustness of tomographic inversion with respect to small changes in model parameterization and regularization. A related benefit is the increased stability of depth images with respect to small changes in velocity, which can increase confidence in interpretation. The velocity models that result from wave tracing match picked arrival times in band-limited data better and generate improved depth images. These advantages of wave tracing relative to conventional Snell’s law ray tracing have been tested on both synthetic and real data examples for crosswell seismic geometry.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have