Icon
  • Home
  • All Lessons
  • About

Variograms

The spatial distribution of some variables is modeled without a variogram. Some categories are modeled by objects or with training images, 3-D solid models of wireframes may be constructed without a kriging-based algorithm and some variables could be mapped deterministically with another technique that does not use a variogram. Yet, most variables in a geostatistical study require a variogram. The variogram may be required of the original variable, the indicator transform of a categorical variable and/or the Gaussian transform of a continuous variable. In general, a variogram is required for every variable in every stationary subset at the coordinate system, scale and transformed units that will be used in modeling. These lessons are concerned with practices in calculating and modeling variograms with uncertainty.

Lessons

  • The Pairwise Relative Variogram (see source code on GitHub)
  • Experimental Variogram Tolerance Parameters
  • Calculation and Modeling of Variogram Anisotropy
  • The Sill of the Variogram
  • Variogram Calculation for Tabular Deposits
  • Transforming a Variogram of Normal Scores to Original Units

  • Locally Varying Anisotropy
  • The Angle Specification for GSLIB Software (see source code on GitHub)
  • The Nugget Effect (see source code on GitHub)

Partners:

Contact Resource Modeling Solutions about a commercial or academic license to RMSP to apply the techniques covered here in Geostatistics Lessons.

© 2015-2025 Jared Deutsch | Terms of Use