A seismic numerical modeling experiment has been conducted to acquire a wide azimuth 3D-3C seismic data from an isotropic heterogeneous elastic model and an anisotropic homogeneous equivalent model in order to verify the suitability of these two modeling approaches for anisotropic studies. This study focuses on reflection amplitude and traveltime comparison of the two models. Although, geophysicists often prefer to use anisotropic homogeneous equivalent models for various seismic modeling and imaging tasks, there are however some benefits of heterogeneous models over anisotropic homogeneous equivalent models. We show that the anisotropic equivalent modeling predicts strong interbed multiples and multimodes which are much weaker in the heterogeneous elastic model. This is because a heterogeneous medium will cause irregular scattering of multiples and multimode events, thus diminishing these events. Both modeling results reveals AVAZ signatures which shows more significant azimuthal variations in the elastic model than in the equivalent model because of the strong multimode conversions which tend to obscure primary reflections at far offsets. Also, we investigated the effect of offset on PP and PS azimuthal anisotropy from the two HTI models with the aim of using the modeling results as guidance in seismic data application. AVAZ analysis shows that the major axes of the radial-component PS-wave amplitude elliptical fit are perpendicular to the fracture strike, which is opposite to the PP-wave amplitude elliptical fit whose major axes are parallel to the fracture strike. The azimuthal interval traveltime shows poor elliptical distribution, making it is difficult to use the interval traveltime analysis for the diagnostic use of fracture orientation. There is an offset limitation for both PP- and PS-waves traveltime fracture-induced anisotropy in both modeling. We observe also that PS-wave amplitudes from both models show a wider applicable offset range and larger observable azimuthal anisotropy than PP-waves.
View full article as PDF (4.90 Mb)