Frequency-domain waveform inversion based on the angle-domain filtered gradient

Y. Choi*, T. Alkhalifah

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Correlating the source and receiver wavefields, which are travelling in a similar direction, isolates the long wavelength features of the gradient. This selective correlation is effectively achieved using the angledomain imaging condition. We propose a frequency-domain waveform inversion algorithm based on the angle-domain filtered gradient. Since a large reflection angle in the imaging condition implies that the source and receiver wavefiels are travelling in a similar direction, the filtered gradient based on the large cut-off angle has a long wavelength component. As the cut-off angle is decreased, the proposed inversion method moves from the tomographic type inversion to the classic waveform inversion. Numerical examples show that the waveform inversion based on the angle-domain filtered gradient generates a long wavelength structure without low frequency information.

Original languageEnglish (US)
Title of host publication77th EAGE Conference and Exhibition 2015
Subtitle of host publicationEarth Science for Energy and Environment
PublisherEuropean Association of Geoscientists and Engineers, EAGE
Pages915-919
Number of pages5
ISBN (Electronic)9781510806627
DOIs
StatePublished - Jan 1 2015
Event77th EAGE Conference and Exhibition 2015: Earth Science for Energy and Environment - Madrid, Spain
Duration: Jun 1 2015Jun 4 2015

Publication series

Name77th EAGE Conference and Exhibition 2015: Earth Science for Energy and Environment

Conference

Conference77th EAGE Conference and Exhibition 2015: Earth Science for Energy and Environment
CountrySpain
CityMadrid
Period06/1/1506/4/15

ASJC Scopus subject areas

  • Geochemistry and Petrology
  • Geophysics

Fingerprint

Dive into the research topics of 'Frequency-domain waveform inversion based on the angle-domain filtered gradient'. Together they form a unique fingerprint.

Cite this