Zhou, Hua-Wei2015-08-282015-08-28May 20152015-05http://hdl.handle.net/10657/1136Salt-body delineation plays a critical role in oil-and-gas exploration in the Gulf of Mexico. Common practices for building a salt model rely on interpreting the top of salt first and then interpreting the base of salt on a migrated image. This process can be extremely time-consuming and resource intensive for a complex salt geometry and can be even more tedious in the presence of salt overhangs. The interpretation of base-salt geometry is often the most challenging part of the workflow, because salt base is typically poorly imaged due to limited illumination, which can be caused by the data acquisition, data quality, aperture, and choice of imaging algorithms. To facilitate the building of salt models, we propose a reflection layer tomography (RLT) that estimates the geometries of velocity interfaces by minimizing the kinematic errors measured on angle domain common image gathers. The goal for reflection layer tomography is to automatically and effectively estimate velocity interfaces such as salt boundary. This approach use common image gathers generated by accurate imaging algorithms such as reverse time migration. To mitigate the non-uniqueness of the tomographic inversion, we have further developed a multi-scale inversion. Compared with the single-scale RLT, the multi-scale RLT delivers superior results in delineating the base-salt geometry, as demonstrated by the synthetic examples. In addition, RLT has been extended to 3D for more practical application. The synthetic results show that both migrated stack and common image gathers can be significantly improved using this approach.application/pdfengThe author of this work is the copyright owner. UH Libraries and the Texas Digital Library have their permission to store and provide access to this work. Further transmission, reproduction, or presentation of this work is prohibited except with permission of the author(s).TomographyMulti-scaleBase-salt geometryGeophysics2015-08-28Thesisborn digital