3D local tomography - Residual interval velocity analysis on a depth solid model

Zvi Koren*, Uri I. Zackhem, Dan Kosloff

*Corresponding author for this work

Research output: Contribution to conferencePaperpeer-review

3 Scopus citations

Abstract

An interactive method for the determination of interval velocities and interface depths in 3D complex areas is presented. Starting with an initial velocity depth model, the tomographic principle is used to relate pertubations in layer slowness and interface depths to traveltime changes along CRP rays (Kosloff et al, 1996). The additional constraint of preserving zero offset traveltimes along normal incidence rays enables us to relate interface depth changes to slowness changes and thus reduce the problem to a single parameter problem. By separating the traveltime error function into two parts, contribution of slowness from the overburden and contribution of slowness from the current location, the global tomography is converted to a simple interval velocity analysis problem. The first step in the proposed method is a special 3D reflector prestack depth migration (Koren et al, 1988). The migration is performed on a coarse output grid, where the output consists of CRP gathers in windows centered around the reflecting horizons (CRP migrated panels) and the corresponding CRP ray paths for the relevant output offsets and azimuths. The second step is velocity-depth model updating. For a given layer, a semblance profile of varying residual interval velocities is calculated along the migration grid. The criteria for picking the residual interval velocities is the flatness of the migrated panels at each location. A map of slowness perturbation is then built. Next, depth perturbations are found for the entire model according to the condition that zerooffset traveltimes remain constant. The velocity-depth model is rebuilt only after the errors at all locations and all layers have been picked. The method handles general complex structures and not only layer-cake type structures. For subsurface representation, we use a solid model that consists of triangulated surfaces surrounding closed volumes. The solid model is designed for handling nonlayer cake structures such as salt bodies, lenses, or truncated formations. The method is demonstrated on a 2D synthetic example and on a 3D field dataset from the Gulf of Mexico.

Original languageEnglish
StatePublished - 1999
Event1999 Society of Exploration Geophysicists Annual Meeting, SEG 1999 - Houston, United States
Duration: 31 Oct 19995 Nov 1999

Conference

Conference1999 Society of Exploration Geophysicists Annual Meeting, SEG 1999
Country/TerritoryUnited States
CityHouston
Period31/10/995/11/99

Fingerprint

Dive into the research topics of '3D local tomography - Residual interval velocity analysis on a depth solid model'. Together they form a unique fingerprint.

Cite this