RFWI (Reflection Full Waveform Inversion) is the band-limited transcription of horizon-based traveltime Tomography. In the latter, as in all variants of ray-based reflection Tomography, the problems of the velocity update and of the depth repositioning of the reflectors in the newly updated model are coupled. Ignoring this velocity – depth coupling is an overlooked and underlying cause of slow convergence in RFWI. We analyze the coupling of velocity-update and depth-repositioning in RFWI. We propose to account for it with a data-space pre-conditioner applied on the residuals of RFWI. We show that this preconditioning improves the convergence of RFWI.
Presentation Date: Monday, October 12, 2020
Session Start Time: 1:50 PM
Presentation Time: 1:50 PM
Location: Poster Station 8
Presentation Type: Poster