CRS-partial stack and spectral content

Abstract

This work had as a general aim at the application of techniques based on the common-reflection-surface stack (CRS), and a specific aim at trace interpolation using the CRS-partial stack in order to analyse the effect on spatial aliasing to garantee that the total spectrum content be limited to the two main spectral quadrants. For this, two synthetic tests were constructed that stayed close to the paraxila ray theory, and to validate the hipotheses, by deleting traces, reconstruction of section, and the addition of noise. Introduction The quality of reflection seismic data is an important aspect in seismic processing and imaging, and part of its analysis is performed in the spectral domain, starting with the field survey paramenters, the processing, the velocity analysis, the inversion and up the migration. Factors like the inhomogeneities in the subsurface, the presence of fault structures and strong velocity contrasts lead to a decrease of the S/N ratio and to a more complicated work flow to precondition the data for velocity analysis, velocity model building, and other processes (Baykulov and Gajewski, 2007). Regularization of seismograms and filling the gaps in cases of missing data usually are performed using different binning and interpolation techniques, as described, for instance, by Brune et al. (1994), Yilmaz (2000), and Fomel (2003). This research topic is part of the CRS stack method, as described by Muller (1999), Jager (1999) and Mann (2002), to improve both the quality of the pre-stacked and stacked sections as measured by the spectrum content by a consistent interpolation process. According to Muller (1999) and Jager (1999) the CRS stacking surface aproximates the traveltime of seismic reflection data more accurately than NMO/DMO stack; therefore, the application of the CRS stacking surface to produce regularized data can be superior to methods based on the conventional techniques of NMO/DMO and binning/interpolation as described by Brune et al. (1994). CRS stack The CRS stack method simulates a zero-offset (ZO) section from multi-coverage seismic reflection data for 2D media without explicit knowledge of the macro-velocity model. The CRS stack surface, as illustrated in Figure 1, is an operator that approximates the true subsurface reflector by a reflector element that locally has the same curvature as the true reflector. The traveltime t(xm,h) of primary reflection events is described by three parameters, α , RN and RNIP, in the hyperbolic form, as: t(xm,h) = ( t0 +2 sinα0 v0 (xm − x0) 2 +2t0 cos2 α0 v0 ( (xm − x0) RN + h2 RNIP )