Redatumação Kirchhoff por soma única para dados sísmicos de múltipla cobertura

Abstract

We present a new redatuming approach to correct the distortions caused by the rough topography and the weathering layer, which is usually inhomogeneous, in land multiple coverage seismic data. The redatuming method is physically more justified than the commonly used static correction methods, which provide - in indicated situations - non-physical adjustments (field static) by displacing the traces in time, which leads to sub-optimal reflections imaging and inversion. The redatuming procedure offered here is different: it transforms irregular multiple coverage reflection seismic data on an uneven surface in a pre-specified flat datum in a regular configuration of sources and receivers, below the complex speed zone. This is done by adding/stacking the input traces along traveltime curves or Kirchhoff type operators that are calculated by ray tracing. In this way, distorted reflections and diffractions in the input data become more regular (for example, hyperbolic) and, thus, recognizable in the output data. For this to happen, Kirchhoff single-stack redatuming operator needs an approximation of the weathering zone speed. Then, multiple coverage output data on the new flat datum can be imaged and inverted more reliably than with standard static correction methods. This thesis presents a review of the theoretical approach to Kirchhoff redatuming based on a single summation and proposes development of a new redatuming algorithm for multiple-coverage pre-stack seismic data. The validation of this algorithm is done by applying it to controlled seismic data generated from synthetic models representing real geological situations. This thesis also presents a study on the sensitivity of the Kirchhoff single-sum redatuming algorithm to velocity model errors. To this end, we used information from the velocities of the weathering zone and shallow structures obtained by two commonly used refraction traveltime tomography algorithms in seismic (Rayinvr and Refratom) and compared the redatuming done in each case to verify the influence of the velocity model on the redatuming. The results show that the presented redatuming adequately corrects for the effects of topography and weathering zone using more accurate models, producing depth migrated images with correct positioning of the reflectors. Applying the redatuming algorithm to poor quality 2D land seismic data provided interpolated and regularized prestack redatumed data, whose depth migrated images show improved quality and increased signal-to-noise ratio, with enhanced and more continuous reflector.