Elastic seismic inversion with geostatistical information: resolution and application to an AVO class III gas reservoir

Abstract

In this work we describe a method for joint inversion of multi-angle partial stacks under geostatistical prior constraints for the estimation of the parameters of an isotropic elastic medium. The method is based on fitting the calculated and observed seismic data and minimizing the deviations of the elastic model configuration from the prior configuration, which is obtained by statistical characterization of well-log data. The statistical component of the model correlates vertically (in time) the CDP elastic properties, regularizing the model to the well-log spatial statistics. Lateral property correlations are not included in this work. We define an appropriate measure of the resolving power of the seismic data for spatially correlated models, and illustrate the improvement of the elastic properties resolution with the number of angle stacks considered and the incidence angle range. We show results for a field case, corresponding to a class III gas reservoir, where the mass density and shear velocities estimated from the elastic inversion are key information for the identification of the gas bearing sands.