Beam Tomography for Subsurface Model Building and Imaging

Tomographic model building attempts to improve the quality of earth models by inspecting the image summation coherency and adjusting model velocities to correct for summation misalignments. Beam tomography operates using similar alignment principles. However, instead of using the standard measurement of misalignment between offset ranges, beam tomography measures the misalignment between individual beams and the image formed by summation of all the beams. Just as for standard tomography, the measurements of beam misalignments are often ambiguous because of noise and cycle skips. Our beam tomography approaches this difficulty by using a Markov-chain, Monte Carlo, global optimization method. Instead of representing the misalignment of a beam by a single shift value, a cross-correlation between a beam and the image is computed over a range of shifts. Earth-model velocities are then updated by using tomographic back-projection and a simulated annealing process to optimize the image coherency. In addition to its data-driven modeling building capability, our Beam Tomography is able to incorporate interpretive information in the tomography process as well. In this abstract, we shall demonstrate the effectiveness of the Beam Tomography technology with synthetic and field examples.