The conventional inversion method for seismic data collected in areas of lateral velocity variation is performed by velocity estimation followed by reflectivity imaging. Efficient and accurate prestack depth migration plays an important role in both velocity estimation and reflectivity imaging. Plane-wave synthesis imaging is a prestack depth migration that has both efficiency and accuracy. In this thesis I develop a sequential seismic inversion method that utilizes plane-wave synthesis imaging as a major tool in both velocity estimation and angle-dependent reflectivity recovery.

For the velocity estimation, I have developed a new tomographic velocity estimation method that uses plane-wave synthesis imaging. Conventionally, reflection tomography is characterized by means of a measurement of residual moveout (RMO) that reflects traveltime errors along the ray paths where the event has moved. The method proposed in this thesis measures the RMO with the help of reflector-dependent plane-wave synthesis imaging.

The accuracy of the plane-wave synthesis in angle-dependent reflectivity recovery is obtained when the prestack data were acquired without any missing traces. In order to recover missing traces that can be found in conventional prestack data, I have developed an interpolation technique. After finding an optimum velocity spectrum of a common-midpoint gather, the missing traces are filled in by hyperbolic modeling with the optimum velocity spectrum.

The application of the sequential inversion developed in the thesis to numerically-simulated data and real data shows that the velocity estimated is accurate enough to align common-reflection point (CRP) events horizontally, and the angle-dependent reflectivity recovered shows an AVA (amplitude variation with angel) anomaly, as expected.

- Introduction
- Imaging by wavefront synthesis
- Interpolation of missing offset traces
- Tomographic velocity estimation
- Angle-dependent reflectivity recovery
- Field data examples
- Unitarity of the depth extrapolation operator
- Datuming by wavefield depth extrapolation
- REFERENCES
- About this document ...

2/5/2001