Matrix probing: A randomized preconditioner for the wave-equation Hessian

This paper considers the problem of approximating the inverse of the wave-equation Hessian, also called normal operator, in seismology and other types of wave-based imaging. An expansion scheme for the pseudodifferential symbol of the inverse Hessian is set up. The coefficients in this expansion are found via least-squares fitting from a certain number of applications of the normal operator on adequate randomized trial functions built in curvelet space. It is found that the number of parameters that can be fitted increases with the amount of information present in the trial functions, with high probability. Once an approximate inverse Hessian is available, application to an image of the model can be done in very low complexity. Numerical experiments show that randomized operator fitting offers a compelling preconditioner for the linearized seismic inversion problem. Acknowledgments. LD would like to thank Rami Nammour and William Symes for introducing him to their work. LD, PDL, and NB are supported by a grant from Total SA. 1

[1]  Laurent Demanet,et al.  Fast Discrete Curvelet Transforms , 2006, Multiscale Model. Simul..

[2]  T. V. Bugueva,et al.  A linearized inverse problem for the wave equation in a sphere , 1996 .

[3]  A. Guitton.  Amplitude and kinematic corrections of migrated images for nonunitary imaging operators , 2004 .

[4]  Gregory Beylkin,et al.  Imaging of discontinuities in the inverse scattering problem by inversion of a causal generalized Radon transform , 1985 .

[5]  Ulf Grenander,et al.  The Distribution of Quadratic Forms in Normal Variates: A Small Sample Theory with Applications to Spectral Analysis , 1959 .

[6]  François Treves,et al.  Introduction to Pseudodifferential and Fourier Integral Operators , 1980 .

[7]  Lexing Ying,et al.  The phase flow method , 2006, J. Comput. Phys..

[8]  Peyman P. Moghaddam,et al.  Curvelet-based migration preconditioning and scaling, GEOPHYSICS, 74, no. 4, A41–A46. , 2009 .

[9]  Felix J. Herrmann,et al.  Multifractional splines: application to seismic imaging , 2003, SPIE Optics + Photonics.

[10]  William W. Symes,et al.  Approximate linearized inversion by optimal scaling of prestack depth migration , 2008 .

[11]  V. Rokhlin,et al.  A randomized algorithm for the approximation of matrices , 2006 .

[12]  V. Rokhlin,et al.  A fast randomized algorithm for the approximation of matrices ✩ , 2007 .

[13]  T. Chan,et al.  Interface preconditionings for domain-decomposed convection-diffusion operators , 1990 .

[14]  Per-Gunnar Martinsson,et al.  Randomized algorithms for the low-rank approximation of matrices , 2007, Proceedings of the National Academy of Sciences.

[15]  Lexing Ying,et al.  3D discrete curvelet transform , 2005, SPIE Optics + Photonics.

[16]  Nathan Halko,et al.  Finding Structure with Randomness: Probabilistic Algorithms for Constructing Approximate Matrix Decompositions , 2009, SIAM Rev..

[17]  Christiaan C. Stolk,et al.  Microlocal analysis of a seismic linearized inverse problem , 2000 .

[18]  E. Candès,et al.  The curvelet representation of wave propagators is optimally sparse , 2004, math/0407210.

[19]  A. P. E. ten Kroode,et al.  A microlocal analysis of migration , 1998 .

[20]  Rami Nammour.  Approximate Inverse Scattering Using Pseudodifferential Scaling , 2009 .

[21]  Laurent Demanet,et al.  Discrete Symbol Calculus , 2008, SIAM Rev..

[22]  M. Shubin,et al.  ALMOST PERIODIC FUNCTIONS AND PARTIAL DIFFERENTIAL OPERATORS , 1978 .

[23]  Gang Bao,et al.  Computation of Pseudo-Differential Operators , 1996, SIAM J. Sci. Comput..

[24]  I. Bechar,et al.  A Bernstein-type inequality for stochastic processes of quadratic forms of Gaussian variables , 2009, 0909.3595.

[25]  L. Demanet.  Curvelets, Wave Atoms, and Wave Equations , 2006 .

[26]  Rakesh.  A Linearised inverse problem for the wave equation , 1988 .

[27]  L. Demanet,et al.  Wave atoms and sparsity of oscillatory patterns , 2007 .

[28]  David E Keyes,et al.  Fifth International Symposium on Domain Decomposition Methods for Partial Differential Equations , 1992 .

[29]  Tony F. Chan,et al.  An application of the probing technique to the vertex space method in domain decomposition , 1991 .

[30]  W. Symes,et al.  Approximate Constant Density Acoustic Inverse Scattering Using Dip-Dependent Scaling , 2009 .

[31]  William W. Symes,et al.  Global solution of a linearized inverse problem for the wave equation , 1997 .

[32]  Felix J. Herrmann,et al.  Curvelet-based migration preconditioning and scaling , 2009 .

[33]  J. Rickett.  Illumination-based normalization for wave-equation depth migration , 2003 .

[34]  F. Herrmann,et al.  Sparsity- and continuity-promoting seismic image recovery with curvelet frames , 2008 .