Geostatistics

Geostatistics analyzes spatially correlated data, especially geological measurements, to describe spatial patterns and estimate values at unsampled locations.

Variogram: A function describing how data differences vary with separation distance and direction. It quantifies spatial dependence and supplies the central input to many geostatistical estimators.
Random field: A collection of random variables indexed by locations in space or time. Geostatistics models spatial measurements as realizations of a random field.
Mineral resource estimation: The estimation of the quantity and spatial distribution of minerals in a deposit. It is a foundational use of geostatistical estimation in geology and mining.
Inverse distance weighting: A deterministic interpolation method that weights observations by a power of their distance from the prediction location. Unlike kriging, it does not derive weights from an explicit spatial covariance model.
Georges Matheron: A French mathematician who founded modern geostatistics and developed its theory of regionalized variables. His formal framework transformed mining interpolation into a statistical discipline.
Kriging: A family of weighted linear estimators that uses spatial covariance to predict values at unsampled locations. It converts a fitted spatial dependence model into estimates and prediction variances.
Stationarity: A property of a stochastic process whose specified statistical features remain invariant under shifts. Many geostatistical models assume stable means or dependence across the study region.
Reservoir characterization: The description of a subsurface reservoir's geometry and properties from incomplete observations. Spatial models help represent uncertain rock properties between wells.
Spatial regression: Regression methods that account for spatial dependence in observations, predictors, or errors. It often explains covariate effects, while geostatistics emphasizes spatial structure and prediction.
Regionalized variable theory: A framework that treats spatially distributed geological quantities as structured random variables. Matheron's formulation supplied the conceptual basis for modern geostatistics.

Linked from 9 pages