Estudo do método de empilhamento SRC e aplicação em dados sintéticos e reais

Abstract

Seismic stacking is a method designed to simulate zero-offset (ZO) seismic sections from multi-coverage seismic data. The resulting simulated ZO section gives rise to a significant increase of the signal to noise ratio. This method is done by means of seismic processing socalled the common mid point (CMP) method, through makes use of the so-called normal moveout and dip moveout (NMO/DMO) corrections. In this work, a new seismic stacking method called common reflection surface (CRS) is used to simulate ZO seismic sections. The main characteristic of the CRS method is: 1) the use of a more general, multi-parametric stacking operator that approximates the traveltimes of the primary reflections for sources and receiver pairs arbitrarily located on the vicinity of the normal ray. The parameters or attributes of the stacking operator are the radii of curvatures of two hypothetical waves called normal incidence point (NIP) wave and normal (N) wave; as well as the emergency angle of the normal ray. The CRS method assumes that the near-surface velocity is a priori known. To determine the abovementioned parameters appropriately, main for the CRS imaging method, is necessary search strategies that use cases special of the approach of second hyperbolic order of the traveltimes. The presented search strategies are: extended-pragmatic CRS and global-local CRS. To show the efficiency of these strategies they are applied in the synthetic Marmousi and real land datasets of the Amazon palaeozoic basin. As result sections ZO simulated by three different stacking methods (CMP, CRS extended- pragmatic and global-local CRS), to compare the efficiency of these, with relationship I cost her computational and the resolution of the seismic image.