Waveform inversion through MBTT formulation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Reports (Research Report) Year : 1993

Waveform inversion through MBTT formulation

Guy Chavent
  • Function : Author
  • PersonId : 833570
François Clément

Abstract

This work deals with automatic determination of background velocity from full waveform data. For this purpose we have designed, in an earlier work, an original approach based on the MBTT (Migration Based Travel Time) reformulation of the waveform acoustic inversion problem. The goal of this paper is to give the present status of the development of this approach. We first describe how, with this approach, we can handle the multishot data which are required to obtain low frequency information on the background velocities. We present then in detail four steps towards the implementation of this approach : separation of propagator/ reflector unknowns via multiscale analysis, separate imaging of impedance and velocity reflectivity, amplitude preserving migration for the proper imaging of deep reflectors, determination of velocities and impedance from their background values and the two corresponding reflectivities. Each step is illustrated on a synthetic test example. Finally we give numerical evidence that, already in the case of a single shot, the MBTT approach yields a dramatic improvement of the objective function, which becomes convex and looses all parasitic local minima, and we describe the calculation of the gradient of the new objective function.

Domains

Other [cs.OH]
Fichier principal
Vignette du fichier
RR-1839.pdf (1.73 Mo) Télécharger le fichier

Dates and versions

inria-00074833 , version 1 (24-05-2006)

Identifiers

  • HAL Id : inria-00074833 , version 1

Cite

Guy Chavent, François Clément. Waveform inversion through MBTT formulation. [Research Report] RR-1839, INRIA. 1993. ⟨inria-00074833⟩
264 View
138 Download

Share

Gmail Facebook Twitter LinkedIn More