Geostatistics

Abstract

In this mini-review the fundamentals of geostatistics are described, from the basic approach to the semivariogram to spatial interpolation methods, as kriging and cokriging. The most common mathemathical functions in semivariogram modelling are reviewed, with a description of the parameters extracted from these models which are used to quantify the spatial continuity of a spatial phenomenom and to interpolate values at unsampled points. The different behavior of a variable in different space directions (anisotropy), and the most important assumption in geostatistics (stationarity) are also discussed. In the final part, we explore the use of one or several other variables to improve the knowledge and the interpolation of one spatial phenomenom (cross variogram and cokriging). Finally, cross-validation, a jacknife method to check the model assumptions used in the kriging is discussed.