True-reflection imaging by amplitude correction in local angle domain

Abstract

A review is given on the work of true-reflection imaging with amplitude correction in the local angle domain conducted by our group, the Modeling and Imaging Laboratory, University of California at Santa Cruz. In order to relate the image amplitude faithfully to the local reflection (scattering) property, we need to recover the local scattering pattern (local scattering matrix defined in the local angle domain) by removing the propagation and acquisition-aperture effects. An important component of true-reflection imaging is to have a true-amplitude propagator for the imaging. We apply a localized WKBJ correction to the regular one-way propagator, so that the correct geometric spreading can be calculated in generally heterogeneous media. In the true-reflection imaging process, first local image matrices are obtained by decomposing the wave fields into local plane waves (beamlets) and applying the imaging condition in the local angle domain. Then amplitude factors for acquisition aperture correction are derived based on the directional illumination analysis and applied to the local image matrices. The final images can be presented either as a total strength image field, or as common reflectionangle image (CRI) gathers at the image points. The CRI gathers of the true-reflection imaging have the correct angle-dependence of reflection coefficients and therefore can be used for local AVA analysis and local inversion. The total-strength image is obtained by summing up the contributions from all the incident-receiving angle pairs. Its amplitude is proportional to the average reflection coefficient of the local reflector over the reflection angles. The 2D SEG/EAGE salt model is used to demonstrate the validity and features of the method. Numerical results indicate that localized WKBJ correction has limited but noticeable improvement on the image amplitude. However, the overall image quality is greatly improved by the acquisition-aperture corrections. The amplitudes along the steep faults and the baseline in the subsalt region are much more uniformly distributed after the correction and closer to the true reflection coefficients; meanwhile the noises in the subsalt region caused by salt body multiples and migration artifacts are also reduced due to the amplitude corrections in the local angle domain.