Extracting the CRS Parameters: A Comparison of Two Methods

By stacking on supergathers of sources and receivers of arbitrary location around a central point, the Common-Reflection-Surface (CRS) method is able to produce simulated zero-offset sections of significant signal-to-noise ratio. To do that, the method employs a multi-parameter moveout, the generalized hyperbolic moveout, in which the parameters (three in 2D and eight in 3D) are directly estimated from the multicoverage data. Among its various advantages, the CRS method has the drawback of a costly estimation<br>of the CRS parameters, presently carried out by multiparameter semblance analysis applied to the data. Here we compare two different methods to extract the CRS parameters, an automatic local-slope detection (or plane-wave destruction) algorithm, that estimates the parameters at a fraction of the cost, and the conventional CRS method.