عنوان مقاله [English]
The Common Reflection Surface stack (CRS) method, which is a generalized form of Common-Mid-point (CMP) method, not only simulates the Zero-Offset (ZO) stack section with high Signal to Noise Ratio (SNR) but also generates very useful attributes as byproducts. One of the methods that use these attributes is Normal-Incident-Point (NIP) wave tomography inversion. In this method, a smooth velocity model is obtained by iteratively minimizing the difference between a model parameter and data parameter generated by the CRS stack method; i.e. RNIP and alpha (Duveneck, 2004). The CRS stack method in its basic form (Jäger, 1999; Mann et al., 1999; Müller et al., 1998) is unable to handle the conflicting dips. Later on, this problem was addressed and solved to some extent (Mann, 2002). By merging the concept of Dip-Move-Out (DMO) and CRS stack method, the Common Diffraction Surface (CDS) stack method was introduced (Soleimani et al., 2009a). This method addresses the conflicting dip into full extent (Soleimani et al., 2009b). This method, which calculates CDS attributes by applying coherence analysis on pre-stack multi-coverage data set, is computationally very expensive. By performing the concept of kinematic and dynamic ray tracing on a smooth velocity model with a lower accuracy, the CDS attributes are obtained in an efficient and fast manner (Shahsavani et al., 2011). The attribute RNIP, which is produced by CRS stack method by coherence analysis in a data-driven manner, is effected by RN while RCDS≡RNIP is obtained by CDS is independent of RN. Therefore, it is more accurate (Shahsavani, 2011). In this work, we propose using RCDS instead of RNIP obtained from CRS stack method. The CDS method first is applied to the constant velocity model and then NIP tomography inversion is applied with the respect of obtained attributes to achieve a new velocity model. Again, the CDS method is performed on the new velocity model and the new attributes are generated. This process iteratively is repeated until the velocity model does not change significantly. Based on two hypothetical wavefronts and their emergence angle (i.e., Kinematic wave field attributes) (Hubral, 1983), the CRS stack travel time is obtained using different concepts (Höcht et al., 1999; Schleicher et al., 1993; Tygel et al., 1997) as follows: where t is the travel time of the reflection, v_0 the near-surface velocity, h the half offset, x0 and t0 are the location and traveltime of the selected ZO output sample (x0, t0), respectively, xm is the distance of the midpoint between the shot (S) and receiver location (R) from x0, RN is the radius of the hypothetical normal (N) wave, RNIP is the radius of the hypothetical normal incidence point (NIP)-wave, and α is their common emergence angle.
For a diffractor at the depth RNIP is equal to RN so-called RCDS. So, Eq. (1) is simplified to: