Kriging

Under Kriging (or: Krigen ) refers to a geostatistical method by which one can interpolate values ​​at locations for which no sample by surrounding measured values ​​or approach.

The South African mining engineer Daniel Krige (1951 ) attempted an optimal interpolation method for the mining industry to develop, based on the spatial dependence of measurement points. The process was later named after him. The French mathematician Georges Matheron (1963 ) developed the " theory of regionalized variables", which provides the theoretical framework developed by Daniel Krige method.

The main advantage over simpler methods such as inverse distance weighting is to consider the spatial variance, which can be determined using the semivariograms. While the weights of the inputs used in the calculation of measured values ​​are determined so that the estimation error variance is minimized for a required value. The error depends on the quality of the variogram and the variogram function.

In simpler interpolation problems can occur with accumulation of measurement points. This is avoided when kriging and through the use of the statistical distances between the influent in the calculation of a point neighbors. Weighted means are thus optimized so that the estimator calculates the true value. Kriging is therefore based on efficient and unbiased estimators. If at one point a clustering on the weights of the points will be reduced within this cluster.

Indicator kriging

A method of kriging is the so-called " indicator kriging ". Here are the data before calculating the semivariogram corresponding to a limit on (0, 1 ) transformed. The result can be interpreted as the probability that the selected threshold is exceeded.