2-D Deformable-layer Tomostatics in Sichuan, China

Journal Title
Journal ISSN
Volume Title

Static correction for near-surface effects is a critical issue for onshore seismic data processing due to its significant impact on imaging the subsurface structure, especially for area with severe topographic and near-surface velocity variations. The key idea to determine static correction is to build an accurate near-surface velocity model, which leads to several methods such as refraction statics, uphole surveys and tomostatics. Among these methods, tomostatics, which builds the near-surface velocity models using tomography, is a promising method. However, in complex near-surface areas, traditional grid tomography is often unable to determine the static correction. This is mainly because the conflict between the need of smaller cell to describe the severe velocity variation and the increasing number of inversion unknowns which leads to solution's uncertainty.

The deformable-layer tomography (DLT) determines the complex near-surface velocity models by inverting for depth-varying velocity interfaces. Both synthetic and field data offer many cases illustrating DLT's effectiveness. The main advantage of DLT over grid tomography is that DLT builds a geologically reasonable model with less inversion unknowns, and can resolve the velocity model better with some constrains such as the result of uphole surveys, which is available in my study. Also, a reversed-velocity interface, which is common in mountainous area and has severe effect on near-surface imaging, may be better solved by DLT.

The survey area of my thesis is in the western Sichuan, China, which is mountainous and has a complex near-surface situation. Thus, I have been motivated to find whether the DLT can be a good solution to this problem. I use the DLT to build the near-surface velocity model and determine the static correction for the area. The final velocity model produced by DLT holds close velocity-depth information compared to uphole survey, and static correction calculated from such model shows improvement, such as an increasing level of reflection coherency on stack section.

Static correction