1887

Abstract

We present a generalised Bayesian framework for joint inversion in high-resolution stratigraphic grids with geostatistical spatial constraints. Data types that can be used in the inversion include, but are not limited to, seismic reflection amplitudes, RMS and interval velocities, well logs and 4D time shifts. Geostatistical inversion is frequently used to generate stochastic model sets from seismic reflection data, but the same spatial constraints can also be applied in deterministic inversion, using various types of data, and to models of any continuous subsurface parameter. Deterministic inversion is a prerequisite of stochastic algorithms based on sequential Gaussian simulation (SGS), which may otherwise risk bias from the choice of random path. The two complementary schemes must be compatible, and our probabilistic, Bayesian formulation can be used in both types of approach. Two real examples are given to illustrate the deterministic geostatistical inversion scheme: an elastic AVA inversion to determine acoustic impedance and Poisson's ratio; and a time-lapse inversion to determine reservoir time strain between base and monitor surveys. The algorithm is very efficient, easy to parallelise, and operates on the full 3D model without partitioning. Runtimes on modern workstations are of the order of a few minutes for typical 3D studies.

Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.20149359
2011-05-23
2024-03-28
Loading full text...

Full text loading...

http://instance.metastore.ingenta.com/content/papers/10.3997/2214-4609.20149359
Loading
This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error