Uncertainty estimation for full waveform inversion with a prior information from depth migration

Hejun Zhu, Siwei Li, Sergey Fomel, Georg Stadler, Omar Ghattas

Research output: Contribution to journalArticle

Abstract

We use Bayesian inference framework to estimate model uncertainties associated with full waveform inversion. A prior covariance operator is built based on the estimated slopes of migrated images and plane-wave construction. Frequencydomain full waveform inversion is used to search for the maximum a posterior model. Evaluation of posterior covariance requires knowledge on the prior-preconditioned Hessian. In this study, we apply a randomized SVD approach to analyze the spectrum of the preconditioned Hessian. Strong decay of its eigenvalues indicates that data are most informative to low dimensional subspaces of model parameters. 2D Marmousi model is used as a numerical example to validate the proposed framework. Comparing random samples from the prior and posterior distributions allows us to estimate model uncertainties associated with full waveform inversion.

Original languageEnglish (US)
Pages (from-to)1409-1414
Number of pages6
JournalSEG Technical Program Expanded Abstracts
Volume34
DOIs
StatePublished - 2015

Fingerprint

waveforms
inversions
eigenvalue
Singular value decomposition
estimates
inference
plane waves
eigenvalues
inversion
Uncertainty
slopes
operators
evaluation
decay

ASJC Scopus subject areas

  • Geotechnical Engineering and Engineering Geology
  • Geophysics

Cite this

Uncertainty estimation for full waveform inversion with a prior information from depth migration. / Zhu, Hejun; Li, Siwei; Fomel, Sergey; Stadler, Georg; Ghattas, Omar.

In: SEG Technical Program Expanded Abstracts, Vol. 34, 2015, p. 1409-1414.

Research output: Contribution to journalArticle

@article{0ab465b86b2c4a0a8f95eba658e58977,
title = "Uncertainty estimation for full waveform inversion with a prior information from depth migration",
abstract = "We use Bayesian inference framework to estimate model uncertainties associated with full waveform inversion. A prior covariance operator is built based on the estimated slopes of migrated images and plane-wave construction. Frequencydomain full waveform inversion is used to search for the maximum a posterior model. Evaluation of posterior covariance requires knowledge on the prior-preconditioned Hessian. In this study, we apply a randomized SVD approach to analyze the spectrum of the preconditioned Hessian. Strong decay of its eigenvalues indicates that data are most informative to low dimensional subspaces of model parameters. 2D Marmousi model is used as a numerical example to validate the proposed framework. Comparing random samples from the prior and posterior distributions allows us to estimate model uncertainties associated with full waveform inversion.",
author = "Hejun Zhu and Siwei Li and Sergey Fomel and Georg Stadler and Omar Ghattas",
year = "2015",
doi = "10.1190/segam2015-5834647.1",
language = "English (US)",
volume = "34",
pages = "1409--1414",
journal = "SEG Technical Program Expanded Abstracts",
issn = "1052-3812",
publisher = "Society of Exploration Geophysicists",

}

TY - JOUR

T1 - Uncertainty estimation for full waveform inversion with a prior information from depth migration

AU - Zhu, Hejun

AU - Li, Siwei

AU - Fomel, Sergey

AU - Stadler, Georg

AU - Ghattas, Omar

PY - 2015

Y1 - 2015

N2 - We use Bayesian inference framework to estimate model uncertainties associated with full waveform inversion. A prior covariance operator is built based on the estimated slopes of migrated images and plane-wave construction. Frequencydomain full waveform inversion is used to search for the maximum a posterior model. Evaluation of posterior covariance requires knowledge on the prior-preconditioned Hessian. In this study, we apply a randomized SVD approach to analyze the spectrum of the preconditioned Hessian. Strong decay of its eigenvalues indicates that data are most informative to low dimensional subspaces of model parameters. 2D Marmousi model is used as a numerical example to validate the proposed framework. Comparing random samples from the prior and posterior distributions allows us to estimate model uncertainties associated with full waveform inversion.

AB - We use Bayesian inference framework to estimate model uncertainties associated with full waveform inversion. A prior covariance operator is built based on the estimated slopes of migrated images and plane-wave construction. Frequencydomain full waveform inversion is used to search for the maximum a posterior model. Evaluation of posterior covariance requires knowledge on the prior-preconditioned Hessian. In this study, we apply a randomized SVD approach to analyze the spectrum of the preconditioned Hessian. Strong decay of its eigenvalues indicates that data are most informative to low dimensional subspaces of model parameters. 2D Marmousi model is used as a numerical example to validate the proposed framework. Comparing random samples from the prior and posterior distributions allows us to estimate model uncertainties associated with full waveform inversion.

UR - http://www.scopus.com/inward/record.url?scp=85018985753&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018985753&partnerID=8YFLogxK

U2 - 10.1190/segam2015-5834647.1

DO - 10.1190/segam2015-5834647.1

M3 - Article

AN - SCOPUS:85018985753

VL - 34

SP - 1409

EP - 1414

JO - SEG Technical Program Expanded Abstracts

JF - SEG Technical Program Expanded Abstracts

SN - 1052-3812

ER -