Quick Links


Robust Iterative Migration Velocity Analysis - Benefits for Imaging with Primaries and with First-order Surface MultiplesNormal access

Authors: E. Cocher, H. Chauris and R.-É. Plessix
Event name: 79th EAGE Conference and Exhibition 2017
Session: Velocity Model Estimation - Theory I
Publication date: 12 June 2017
DOI: 10.3997/2214-4609.201700601
Organisations: EAGE
Language: English
Info: Extended abstract, PDF ( 629.39Kb )
Price: € 20

Migration Velocity Analysis techniques suffer from the presence of artefacts in reflectivity models obtained by standard migration. To overcome this difficulty, recent studies suggest determining the reflectivity by inversion instead of migration, either through pseudo-inverse formulas or by iterative minimisation of a least-squares data misfit. The advantage of the iterative approach is the possible introduction of multiple reflections in MVA techniques, usually restricted to primary reflection data. In this nested optimisation process, we pay attention at the computation of the gradient with respect to the background velocity model. From a numerical point of view, we show that this gradient appears to be unstable: in particular, gradients computed with very close reflectivity models may exhibit significant differences. To obtain a more stable procedure, we introduce a slight modification in the MVA objective function. The robustness of the approach is illustrated on simple 2D models, first in the case of primaries only, then with primaries and first-order surface-related multiples.

Back to the article list