Source-independent elastic waveform inversion using a logarithmic wavefield

Handle URI:
http://hdl.handle.net/10754/562039
Title:
Source-independent elastic waveform inversion using a logarithmic wavefield
Authors:
Choi, Yun Seok; Min, Dong Joon
Abstract:
The logarithmic waveform inversion has been widely developed and applied to some synthetic and real data. In most logarithmic waveform inversion algorithms, the subsurface velocities are updated along with the source estimation. To avoid estimating the source wavelet in the logarithmic waveform inversion, we developed a source-independent logarithmic waveform inversion algorithm. In this inversion algorithm, we first normalize the wavefields with the reference wavefield to remove the source wavelet, and then take the logarithm of the normalized wavefields. Based on the properties of the logarithm, we define three types of misfit functions using the following methods: combination of amplitude and phase, amplitude-only, and phase-only. In the inversion, the gradient is computed using the back-propagation formula without directly calculating the Jacobian matrix. We apply our algorithm to noise-free and noise-added synthetic data generated for the modified version of elastic Marmousi2 model, and compare the results with those of the source-estimation logarithmic waveform inversion. For the noise-free data, the source-independent algorithms yield velocity models close to true velocity models. For random-noise data, the source-estimation logarithmic waveform inversion yields better results than the source-independent method, whereas for coherent-noise data, the results are reversed. Numerical results show that the source-independent and source-estimation logarithmic waveform inversion methods have their own merits for random- and coherent-noise data. © 2011.
KAUST Department:
Physical Sciences and Engineering (PSE) Division
Publisher:
Elsevier
Journal:
Journal of Applied Geophysics
Issue Date:
Jan-2012
DOI:
10.1016/j.jappgeo.2011.10.013
Type:
Article
ISSN:
09269851
Sponsors:
This work was financially supported by "Development of Technology for CO2 Marine Geological Storage" program of the Ministry of Land, Transport and Maritime Affairs (MLTM), and the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2010-0006155), and the Energy Efficiency & Resources of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Knowledge Economy (No. 2010T100200133). The first author would like to thank Tariq A. Alkhalifah for his helpful review and discussion. The first author would also like to thank the King Abdullah University of Science and Technology for financial assistance in this research. The authors thank Christos Saragiotis for English editing.
Appears in Collections:
Articles; Physical Sciences and Engineering (PSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorChoi, Yun Seoken
dc.contributor.authorMin, Dong Joonen
dc.date.accessioned2015-08-03T09:43:22Zen
dc.date.available2015-08-03T09:43:22Zen
dc.date.issued2012-01en
dc.identifier.issn09269851en
dc.identifier.doi10.1016/j.jappgeo.2011.10.013en
dc.identifier.urihttp://hdl.handle.net/10754/562039en
dc.description.abstractThe logarithmic waveform inversion has been widely developed and applied to some synthetic and real data. In most logarithmic waveform inversion algorithms, the subsurface velocities are updated along with the source estimation. To avoid estimating the source wavelet in the logarithmic waveform inversion, we developed a source-independent logarithmic waveform inversion algorithm. In this inversion algorithm, we first normalize the wavefields with the reference wavefield to remove the source wavelet, and then take the logarithm of the normalized wavefields. Based on the properties of the logarithm, we define three types of misfit functions using the following methods: combination of amplitude and phase, amplitude-only, and phase-only. In the inversion, the gradient is computed using the back-propagation formula without directly calculating the Jacobian matrix. We apply our algorithm to noise-free and noise-added synthetic data generated for the modified version of elastic Marmousi2 model, and compare the results with those of the source-estimation logarithmic waveform inversion. For the noise-free data, the source-independent algorithms yield velocity models close to true velocity models. For random-noise data, the source-estimation logarithmic waveform inversion yields better results than the source-independent method, whereas for coherent-noise data, the results are reversed. Numerical results show that the source-independent and source-estimation logarithmic waveform inversion methods have their own merits for random- and coherent-noise data. © 2011.en
dc.description.sponsorshipThis work was financially supported by "Development of Technology for CO2 Marine Geological Storage" program of the Ministry of Land, Transport and Maritime Affairs (MLTM), and the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2010-0006155), and the Energy Efficiency & Resources of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Knowledge Economy (No. 2010T100200133). The first author would like to thank Tariq A. Alkhalifah for his helpful review and discussion. The first author would also like to thank the King Abdullah University of Science and Technology for financial assistance in this research. The authors thank Christos Saragiotis for English editing.en
dc.publisherElsevieren
dc.subjectBack-propagationen
dc.subjectLogarithmic wavefielden
dc.subjectSource-estimationen
dc.subjectSource-independenten
dc.subjectWaveform inversionen
dc.titleSource-independent elastic waveform inversion using a logarithmic wavefielden
dc.typeArticleen
dc.contributor.departmentPhysical Sciences and Engineering (PSE) Divisionen
dc.identifier.journalJournal of Applied Geophysicsen
dc.contributor.institutionDepartment of Energy Systems Engineering, Seoul National University, 599 Gwanak-ro, Gwanak-gu, Seoul, 151-744, South Koreaen
kaust.authorChoi, Yun Seoken
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.