1. Migration Deconvolution
    • RESULT: Successful test on VSP synthetic and field data. MD makes the VSP migration look more like a CDP migration image. Performance of MD varies depending on severity of acquisition footprint. MD processing works better on Kirchhoff image compared to wave equation image.
    • IMPLICATION: Because of the odd acquisition footprint associated with VSP images we conclude that MD should be applied to most VSP migration images.
    • PROBLEMS: Attention should be paid to preserving low wavenumber components in image.
    • 2005 GOALS: Assign student to make codes useable. Extrapolation deconvolution.

  2. Multiple Prediction and Migration
    • RESULT: Several approaches. 1). The extended Wiggin's extrapolation-prediction method is now carried out with an efficient phase-shift extrapolation method. The 3D version of this method is successfully tested for a simple synthetic model. 2). Interferometric migration of multiples.
    • IMPLICATION: Several new approaches to either eliminating multiples or imaging multiples.
    • PROBLEMS: Need further field data tests. Did not get to 3D VSP data because the data we have has insufficient quality.
    • 2005 GOALS: Full 3-D tests with CDP field data. Multiple migration of 3D VSP data.

  3. Seismic Interferometry
    • RESULT: Several new results: 1). Interferometric extrapolation, 2). Interferometric Fermat's principle, 3). Interferometric imaging of PS waves and salt flanks, 4). Interferometric traveltime tomography for target oriented imaging and tomostatics using multiples, 5). Relationship between interferometry and virtual source imaging, 6). Reduced time migration vs interferometric migration, 7). Interferometric deconvolution of near-surface scattering.
    • IMPLICATIONS: Opens up a new field in seismic imaging. Will try to solve tough practical problems.
    • PROBLEMS: Immature testing phase for field data. Interferometric imaging is an order-of-magnitude more expensive than standard migration. How sensitive is II to wrong choice of reference layer?
    • 2005 GOALS: Test above methods on tougher data sets.

  4. Primary Only Imaging Condition + Radon Filtering of Multiples
    • RESULT: POIC+Radon filtering of SMAART data multiples. Extrapolation and filtering strategy works on synthetic data.
    • IMPLICATIONS: Can eliminate some near-offset multiples.
    • PROBLEMS: Not a perfect multiple killer, but can be a sueful tool in arsenal of multiple methods.
    • 2005 GOALS: Done.

  5. First Arrival Waveform Imaging and Refraction Migration
    • RESULT: Several results. 1). Waveform inversion test on BP synthetic shows big improvement in migration image compared to migration with traveltime tomogram. 2). INCO crosswell data shows waveform tomogram is only a slight improvement over traveltime tomogram in delineating ore boundary. 3). Waveform tomograhy initiated for Saudi data.
    • IMPLICATION: Synthetic data results are encouraging.
    • PROBLEMS: BP data very time consumming.
    • 2005 GOALS: Full tests on Saudi, Inco data and other data. Elastic/viscoelastic inversion.