3D Stratigraphic models : representation and stochastic modeling

in: Proc. IAMG 2006

Abstract

Several representations have been proposed to model the 3D architecture and petrophysical properties of stratigraphic formations from subsurface data. All methods aim at finding a consistent way of describing heterogeneities by defining a curvilinear coordinate system conform to the sedimentary layers. Classically, this coordinate system is implicitly defined by a grid conforming to the strata and to the faults; alternatively, the Geo-chronological model (Geochron) explicitly defines this coordinate system, without using such a grid. This paper compares the grid-based stratigraphic modeling methods with the Geochron method. Their ability to represent complex fault networks is discussed, proving the higher representative power of the Geochron model. Because the visualization of uncertainty calls for generating several possible images of the subsurface, we also discuss grid perturbation methods, and introduce a new method to stochastically perturb a Geochron model.

Download / Links

BibTeX Reference

@INPROCEEDINGS{Caumon06IAMGb,
    author = { Caumon, Guillaume and Mallet, Jean-Laurent },
     title = { 3D Stratigraphic models : representation and stochastic modeling },
 booktitle = { Proc. IAMG 2006 },
      year = { 2006 },
  abstract = { Several representations have been proposed to model the 3D architecture and petrophysical properties of stratigraphic formations from subsurface data. All methods aim at finding a consistent way of describing heterogeneities by defining a curvilinear coordinate system conform to the sedimentary layers. Classically, this coordinate system is implicitly defined by a grid conforming to the strata and to the faults; alternatively, the Geo-chronological model (Geochron) explicitly defines this coordinate system, without using such a grid. This paper compares the grid-based stratigraphic modeling methods with the Geochron method. Their ability to represent complex fault networks is discussed, proving the higher representative power of the Geochron model. Because the visualization of uncertainty calls for generating several possible images of the subsurface, we also discuss grid perturbation methods, and introduce a new method to stochastically perturb a Geochron model. }
}