Seismic Attributes for Constraining Geostatistical Seismic Inversion

This paper aims to constrain the geostatistical seismic inversion method, improving it to match acoustic and/or elastic models with spatial structures interpreted from seismic attribute analysis. A method using seismic attributes as parameters in the objective function was created and inserted within the standard Global Seismic Inversion approach, where the a global perturbation method is done using Direct Sequential Simulation and Co-Simulation as the image transforming technique. Convergence is measured by comparing selected seismic attributes calculated from the synthetic seismic data with those derived from the real seismic dataset. The algorithm was tested on a real case study from a deep- water carbonate oil reservoir. Several combinations of seismic attributes were tested to determine the method's sensibility. The approach presented here can be used to constrain the inherent spatial uncertainty, associated with geostatistical seismic inversion processes with features that are inferred from the seismic signal, the seismic attributes.