Fast Geostatistical Realization
Date: December 19, 2019
The next release of Earth Volumetric Studio will also include our Patent Pending Fast Geostatistical Realization™ (FGR™) technology. This is the “secret” that we could not talk about prior to filing the patents.
Our Geostatistical Realizations are a significant advancement to industry standard Gaussian Geostatistical Simulations (GGS). A good reference on GGS can be found on ESRI’s website.
Traditional GGS provides realizations (potential scenarios) showing how a geostatistical model might vary from the nominal prediction due to the computed standard deviations at every point on the grid. The methodology used to create GGS results in a computationally intensive process that makes the creation of a single realization a very slow process taking far longer than the creation of the nominal model. Furthermore, this approach creates variations that fluctuate at a wavelength that is the grid spacing. This creates very noisy results that do not realistically model the physical world.
C Tech’s FGR™ technology was developed to address the shortcomings of traditional GGS, while providing all of its benefits across the full range of modeling that we perform. With FGR™ you can:
- Create realizations of 2D & 3D analytical data
- Create realizations of stratigraphic models (horizons) using kriging or Natural Neighbor
- Create realizations of 3D smooth and block lithologic models
- Create all realizations in a fraction of the time required to create the Nominal Models.
- Control the wavelength & magnitudes of statistical variations
- Apply simple or fully 3D anisotropy to the magnitudes of statistical variations
FGR™ technology provides an important additional tool beyond C Tech’s Min/Max (plume) technology for many reasons:
- Min/Max is only available for analytical modeling in 2D & 3D
- The Min and Max scenarios are a worst case applied over the full model extents
- It is statistically unlikely that real-world site conditions (realizations) will be everywhere at the minimum or maximum.
- With FGR™ you can create a multitude of more real-world site conditions
- FGR™ variations can be limited by statistical confidence just like Min/Max
- FGR™ can be applied to all types of modeling