Introducing students to theoretical concepts and basic concepts in the field of geostatistics and spatial statistical methods. Practical application of geostatistical methods through current software packages.

Each student should be able to model and numerically represent the structure of the spatial correlation of the observed phenomenon, measured at discrete points in the domain of interpolation, to select and apply an adequate method to perform spatial prediction of the observed magnitude at points where no observations were made.

Introduction to Geostatistics. The concept of spatial modeling. Regionalized variable theory. Types of spatial variables. Review and division of interpolation methods.

Spatial prediction and interpolation. Surface trend regression models. Multiple regression.

Characteristics of spatial variability. The notion of variograms and covariance. Spatial covariance, stationarity and ergodicity. Experimental variogram and experimental covariance function.

Modeling of variogram function by least squares method. The concept of anisotropy.

Theory: ordinary kriging, universal kriging, initiator kriging, co-kriging, block kriging.

Prediction quality prediction of geostatistical methods. Kriging variance. Cross-validation. Quality assessment using an independent data set. Stochastic simulations.

Visualization of prediction and uncertainty.

Burrough P., McDonnell R.: Principi geografskih informacionih sistema (srpski prevod), Građevinski fakultet, Beograd, 2006.

Armstrong M.: Basic linear geostatistics, Springer, 1998