Seismic data interpolation using plane-wave shaping regularization

Conclusions

Plane-wave shaping (PWS) is a powerful tool for constraining the solutions to inverse problems that require conforming to a local plane-wave structure. The impulse response in both 2-D and 3-D verify this image-guided nature. We demonstrate its effectiveness as an interpolation scheme using simple missing-data interpolation experiments on both real and synthetic examples and show that the method converges to a low misfit in fewer iterations than alternative regularization schemes which use either PWD or PWC. The benefit of fast convergence comes from the fact that oftentimes large-scale geophysical inverse problems can only afford a small number of iterations. When stopping the inversion after only a small number of iterations, PWS produces a more accurate estimation of model parameters than PWD and PWC. This fact gives PWS high potential for geophysical applications beyond simple interpolation.