Wide azimuth tomography ‐ is it necessary?

Abstract

The goal of seismic tomography is to estimate a velocity model, so that migration produces an optimum image. Velocity models derived from seismic data using tomography suffer from non‐uniqueness, but in general, the resolution of the model depends on the measurements used and the parameterization of the model. Unfortunately, the more measurements we use in the tomography, and the greater degree of freedom we allow in the model, the costlier it becomes to accomplish the task. Our goal is to balance the demands of optimizing the velocity model against the costs of picking, QC'ing and inverting the measurements. The data acquired during the Wide Azimuth Towed Streamer (WATS) field trial over the Mad Dog field in the Gulf of Mexico provided a unique opportunity to compare results from three different tomography experiments. In the first experiment we ran tomography based on single parameter residual moveout using a narrow azimuth subset of our wide azimuth data. In the second experiment we used the same data, but used tomography based on multiparameter residual moveout. As expected this second velocity model contained higher resolution details than the first. In the final experiment we used the entire wide azimuth dataset applying tomography based on multiparameter vector‐offset residual moveout. Our study showed that using all of the data provided a better velocity model in the areas with complicated structures than either of the approaches above. Thus, although costly, the tomography that incorporates all of the data provides a higher resolution velocity model that should be used to obtain the best imaging results.