
Full text loading...
The majority of the geostatistical models used in the petroleum industry are variogram-based models. The variogram enables to build effective estimation (kriging) and simulation operators by catching the mean spatial correlation inherent to a data set. Despite their popularity, variogram-based models are the subject of several criticisms. The main one is that they should not be able to handle and reproduce precisely the structural complexity of the reality. M-GS models are improved variogram-based models. They are based on an optimized determination of structural and computational parameters. They lead to more precise and/or more realistic estimation and simulation results than conventional variogram-based models. It is shown by four examples regarding gridding (2), noise removal and simulation issues.