Registration-Based Guided Least-Squares Waveform Inversion

Full waveform inversion with frequency sweeping cannot start from zero frequency because of the lack of low-frequency data, requiring a good starting model. We study a different iterative scheme where the notion of proximity of two traces is not the usual least-squares distance, but instead involve...

Full description

Bibliographic Details
Main Authors: Calandra, Henri, Baek, Hyoung Su, Demanet, Laurent
Other Authors: Massachusetts Institute of Technology. Department of Mathematics
Format: Article
Language:en_US
Published: European Association of Geoscientists & Engineers 2017
Online Access:http://hdl.handle.net/1721.1/109837
https://orcid.org/0000-0001-7052-5097
_version_ 1826210793801121792
author Calandra, Henri
Baek, Hyoung Su
Demanet, Laurent
author2 Massachusetts Institute of Technology. Department of Mathematics
author_facet Massachusetts Institute of Technology. Department of Mathematics
Calandra, Henri
Baek, Hyoung Su
Demanet, Laurent
author_sort Calandra, Henri
collection MIT
description Full waveform inversion with frequency sweeping cannot start from zero frequency because of the lack of low-frequency data, requiring a good starting model. We study a different iterative scheme where the notion of proximity of two traces is not the usual least-squares distance, but instead involves registration as in image processing. In order to create transported data, we introduce a nonconvex optimization problem and solve it in a multiscale fashion from low to high frequencies. This process requires defining low-frequency augmented signals in order to seed the frequency sweep at zero frequency. Successful registrations of noisy data, and application of the new method to model velocity estimation are demonstrated. In a crosshole seismic inversion example (transmission setting), we show that the new method decreases the model velocity error while conventional least-squares inversion converges to a spurious model.
first_indexed 2024-09-23T14:55:48Z
format Article
id mit-1721.1/109837
institution Massachusetts Institute of Technology
language en_US
last_indexed 2024-09-23T14:55:48Z
publishDate 2017
publisher European Association of Geoscientists & Engineers
record_format dspace
spelling mit-1721.1/1098372022-09-29T11:29:38Z Registration-Based Guided Least-Squares Waveform Inversion Calandra, Henri Baek, Hyoung Su Demanet, Laurent Massachusetts Institute of Technology. Department of Mathematics Baek, Hyoung Su Demanet, Laurent Full waveform inversion with frequency sweeping cannot start from zero frequency because of the lack of low-frequency data, requiring a good starting model. We study a different iterative scheme where the notion of proximity of two traces is not the usual least-squares distance, but instead involves registration as in image processing. In order to create transported data, we introduce a nonconvex optimization problem and solve it in a multiscale fashion from low to high frequencies. This process requires defining low-frequency augmented signals in order to seed the frequency sweep at zero frequency. Successful registrations of noisy data, and application of the new method to model velocity estimation are demonstrated. In a crosshole seismic inversion example (transmission setting), we show that the new method decreases the model velocity error while conventional least-squares inversion converges to a spurious model. TOTAL (Firm) 2017-06-13T19:26:26Z 2017-06-13T19:26:26Z 2013-06 Article http://purl.org/eprint/type/ConferencePaper 978-90-73834-48-4 http://hdl.handle.net/1721.1/109837 Baek, Hyoungsu, Henri Calandra and Laurent Demant. "Registration-Based Guided Least-Squares Waveform Inversion." 75th EAGE Conference & Exhibition incorporating SPE EUROPEC 2013, 10 June, 2013, London, United Kingdom, EAGE, 2013. https://orcid.org/0000-0001-7052-5097 en_US http://dx.doi.org/10.3997/2214-4609.20130410 Proceedings of the 75th EAGE Conference & Exhibition SPE EUROPEC 2013 Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf European Association of Geoscientists & Engineers MIT web domain
spellingShingle Calandra, Henri
Baek, Hyoung Su
Demanet, Laurent
Registration-Based Guided Least-Squares Waveform Inversion
title Registration-Based Guided Least-Squares Waveform Inversion
title_full Registration-Based Guided Least-Squares Waveform Inversion
title_fullStr Registration-Based Guided Least-Squares Waveform Inversion
title_full_unstemmed Registration-Based Guided Least-Squares Waveform Inversion
title_short Registration-Based Guided Least-Squares Waveform Inversion
title_sort registration based guided least squares waveform inversion
url http://hdl.handle.net/1721.1/109837
https://orcid.org/0000-0001-7052-5097
work_keys_str_mv AT calandrahenri registrationbasedguidedleastsquareswaveforminversion
AT baekhyoungsu registrationbasedguidedleastsquareswaveforminversion
AT demanetlaurent registrationbasedguidedleastsquareswaveforminversion