For 2D and 3D PSDM projects in thrust belt, the velocity field determination needs to be accurate in order to avoid wrong reflector depths and geometry.

Geosystem’s tomographic tools can solve several problems associated to seismic exploration methods by providing robust estimates of Vint/Depth from surface to depth.

The input to tomographic inversions can be pre-stack travel-time data (e.g. First Breaks, Head Waves or Reflection travel-times) or residuals of the post-migrated domains (i.e. Common Image Gather residuals). Joint inversion schemes are also provided.

Both P and S waves can be used for tomographic applications as well as P-S converted modes.

Geosystem has on-going R&D projects aimed at continuously improving the reliability and the efficiency of velocity determinations through inversion, since we know that the accurate estimate of velocities in depth domain is the key factor for successful imaging.

Thrust-Belt 3D Tomography: Layer-based Tomographic Inversion (left); Grid-based Tomographic Inversion (right).

MODEL PARAMETERIZATION (LAYER-based or GRID-based)

Layer-based or Grid-based model parameterizations are selected depending on the litostratigraphic framework. Layer-based tomography is suited for rock formations such as carbonate bodies where the velocity changes occur mainly at the interfaces and the internal velocity gradients are negligible.

On the other hand, there are geologic conditions for which the depth of burial and the age of the sedimentary formations are important factors controlling the velocity other than the lithology. The most appropriate way to model these situations is the use of a grid-based tomographic approach, which allows the definition of lateral and vertical velocity gradients.

A combination of the two model parameterizations is also necessary when a compact unit (like a salt body) is located within a sedimentary sequence. In such cases the best parameterization approach consists of a grid based approach where the sediment section is allowed to assume smooth velocity functions (i.e. gradients) and the salt body is forced to assume a relatively uniform velocity. This is achieved by appropriately setting the model covariance matrix in the inversion.

Ray-tracing 3D (left); Tomographic Inversion of CIG residuals (right).

3D MODEL BUILDING THROUGH INVERSION OF TURNING RAY AND CIG RESIDUALS

The residuals observed on the common image gathers are related to uncorrected velocity fields and the observed residual move-out in depth domain is used to determine the velocity updates to the model by means of a tomographic procedure. The depth residuals are transformed to time residuals, ray-tracing (with bent rays) is performed and the tomographic inversion takes place to update the velocity field. Grid-based or Layer-based inversions can be carried out using a global inversion approach (i.e. inverting for the whole model) or layer by layer (i.e. layer stripping) proceeding from top to bottom of the model.

Grid-based (left) vs Layer-based (right) model parameterizations.

top of page