| Popis: |
The 3D ZOCRS operator is based on a second order approximation of traveltimes in midpoint and half-offset coordinates (Cristini 2003). The stacking surface in the timespace domain, used to compute the CRS volume, depends on eight parameters that describe the shape and direction of normal ray and wavefront curvatures for Normal (N) wave and Normal Incidence Point (NIP) wave. The eight parameters are obtained with only two searches (a three-parameters followed by a five-parameter search) determined by a coherency analysis procedure. The method does not require the exact knowledge of the velocity model, an a priori knowledge of the near surface constant velocity being sufficient. However in presence of multiples it is possible to constrain the parameters’ space using an input velocity field. These boundaries can be obviously relaxed depending on the reliability of the velocity field. Moveout corrections are stretch-free, computed automatically for every sample and not limited to single CMP locations, but are extended, in a physically consistent manner, to super-gather defined on the projected Fresnel zones. This implies a better focusing of the events and a dramatic increase of fold. Adding to the previous ones the contribution of the structural term derived from the N-wavefront leads a superior imaging quality. Moreover, considering this approach as a data driven technique, the auxiliary outputs (e.g. projected Fresnel zones and structural term) may provide useful information for inversion and interpretation. 3D CRS Processing: A Better Use of Pre-stack Data |
