What is it about?
The authors propose a generalized methodology to define a joint distribution of elastic and petrophysical properties, based on a rock-physics model parametrized by facies. The distribution is used as a prior model in a Bayesian framework to jointly estimate the facies, the elastic and the petro-physical properties of the subsurface.
Featured Image
Why is it important?
If we include all the spatial dependencies of the subsurface properties, the computation of the posterior probability of model parameters conditioned by data would be unfeasible. In this work, we propose a Monte Carlo algorithm integrated with geostatistical methods to numerically compute the spatially dependent posterior distribution. The sampling is based on an approximate posterior distribution, without spatial correlation, which can be analytically computed.
Perspectives
We hope this work might introduce the Bayesian approach as an efficient technique for model parameter estimation and uncertainty quantification in inverse problems. We also expect that the article will help geophysicists with the proposed petrophysical inversion method that can be applied in real datasets with a low-computational cost. Besides, we believe the text can be helpful for non-expert professionals in the understanding the basic concepts of the Bayesian approach to solve inverse problems.
Leandro Passos de Figueiredo
LTrace Geophysical Solutions
Read the Original
This page is a summary of: Joint Bayesian inversion based on rock-physics prior modeling for the estimation of spatially correlated reservoir properties, Geophysics, September 2018, Society of Exploration Geophysicists,
DOI: 10.1190/geo2017-0463.1.
You can read the full text:
Contributors
The following have contributed to this page
