In this section I present a new processing sequence designed mainly for 3D wide-azimuth land surveys. However, after organization of land data as common azimuths, land and marine processing should be similar and the techniques described are applicable to either environment. The basic processing flow in Figure flow outlines the major steps for amplitude preserving-processing. Although the focus is on the imaging step, a good deal of care is required in the preprocessing stage to restore amplitudes and to suppress the noise.
 |
The sequence features two wave-equation processes, namely AMO and prestack migration where both algorithms preserve the angle-dependent reflectivity. Given the scope of this work, I only developed a depth-dependent prestack depth migration for better control of amplitude weights. A band-pass time filter, whose maximum frequency is determined by the local slope of the operator, is applied to avoid the aliasing of the Kirchhoff operator along its steep traveltime slopes.
The data is organized as common-azimuth and common-offset cubes using the AMO transformation. To account for the fold variations of the data and the time-space variability of the operator aperture, AMO is calibrated by the response of a flat event. Finally, prestack migration is applied to the regularized data to determine the location and extent of the anomalies. At this stage any efficient migration algorithm which requires regular grid can be applied to the regularized subset. I chose Kirchhoff migration for consistency in comparing the results of imaging before and after regularization.