Fast Least-Squares Migration with Multiples and Source Estimation

Ning Tu, Aleksandr Y. Aravkin, Tristan van Leeuwen, Felix J. Herrmann

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


The advent of modern computing has made it possible to$ndo seismic imaging using least-squares reverse-time$nmigration. We obtain superior images by solving an$noptimization problem that recovers the$ntrue-amplitude images. However, its success hinges$non overcoming several issues, including overwhelming$nproblem size, unknown source wavelet, and$ninterfering coherent events like multiples. In this$nabstract, we reduce the problem size by using ideas$nfrom compressive sensing, and estimate source$nwavelet by generalized variable projection. We also$ndemonstrate how to invert for subsurface information$nencoded in surface-related multiples by$nincorporating the free-surface operator as an areal$nsource in reverse-time migration. Our synthetic$nexamples show that multiples help to improve the$nresolution of the image, as well as remove the$namplitude ambiguity in wavelet estimation.
Original languageEnglish
Title of host publication75th EAGE Conference & Exhibition
PublisherEuropean Association of Geoscientists and Engineers, EAGE
Number of pages4
Publication statusPublished - 2013
Externally publishedYes


  • expanded


Dive into the research topics of 'Fast Least-Squares Migration with Multiples and Source Estimation'. Together they form a unique fingerprint.

Cite this