lithologic realization

lithologic realization

The lithologic realization module is one of three similar modules (the other two are analytical_realization and stratigraphic_realization), which allows you to very quickly generate statistical realizations of your 2D and 3D lithologic models based upon C Tech’s Proprietary Extended Gaussian Geostatistical Simulation (GGS), which we refer to as Fast Geostatistical Realizations^®^ or FGR^®^. Our extensions to GGS allow you to:

  • Create realizations rapidly:
    • Though indicator_realizations are the slowest of the three because:
    • The material probabilities must be additionally processed to assign materials
    • When Smooth option is on, this process often takes nearly as long as the original kriging
  • Exercise greater control over the frequency and magnitude of visual noise typical of GGS.
  • Control deviation magnitudes from the nominal kriged probability prediction based on a Min Max Confidence Equivalent.
    • Deviations are the absolute value of the changes to each material’s probability
  • Apply Simple or Advanced Anisotropy control over 2D or 3D wavelengths

Module Input Ports

  • Realization [Special Field] Accepts outputs from 3d estimation and krig_2d to allow for EGGS models to be created

Module Output Ports

  • Output Field [Field] Outputs the subsetting level
  • Deviations Field [Field] Outputs the deviations from the nominal kriged probabilities

Important Parameters

There are several parameters which affect the realizations. A brief description of each follows:

  • Randomness Generator Type
    • There are four types, each of which create a different 2D/3D random distribution
  • Anisotropy Mode
    • Two options here are Simple or Advanced. These are equivalent to the variogram settings in lithologic modeling
  • Seed
    • The “Seed” is used in the random number generator, and makes it reproducible.
    • Unique seeds create unique realizations
  • Wavelength
    • The 2D or 3D random distribution is governed by a mean wavelength that determines the apparent frequency of deviations from the nominal kriged probabilities results.
    • Wavelength is in your project coordinates (e.g. meters or feet)
    • Longer wavelengths create smoother realizations
    • Shorter wavelengths create more “noisy” variations in the realizations
    • Very short wavelengths will give results more similar to GGS (aka Sequential Gaussian Simulations)
  • Min Max Confidence Equivalent
    • This parameter determines the magnitude of the deviations.
    • Values close to 50% result in outputs that deviate very little from the nominal kriged probabilities results.
      • (we do not allow values below 51% for algorithm stability reasons)
    • Values at or approaching 99.99% will result in the greatest (4 sigma) variations (more similar to GGS)