Abstract
The word kriging, which is a basic concept of geostatistics, was named from D. G. Krige, a South African mining engineer, who developed a new estimation method in the early 1950s. Georges Matheron, who was a French engineer and currently known as a creator of geostatistics, formalized Krige's innovative concepts in the early 1960s. It was the mid-to late 1980s that geostatistical techniques started to be used in the petroleum industry. The reasons why geostatistics has been spread so much in the petroleum industry are the three major features as follows, 1) Geostatistical techniques are mapping (or modeling) tools utilizing the spatial continuity of existing data. 2) They can estimate uncertainty from variance distribution. 3) They can integrate geology, geophysics, and engineering data.
For example, kriging not only provides a spatial distribution but also estimate error-variance at any unknown data point. Kriged results are depended on a variogram model, which has been build from known data (e. g. well-log measurement). Cokriging, expanded from kriging, uses different kind of data. When porosity value should be estimated, the related acoustic data, such as seismic data, are combined. Sequential simulation (or cosimulation) techniques, based on kriging (or cokriging), can provide the equi-probable images more heterogeneous than kriged (or cokriged) results. They are now indispensable tools of reservoir characterization.
