Application of Bayesian Kriging to subsurface characterization

Abstract

A large number of geologic environments are composed of different layers of contrasting grain sizes. An accurate description of the locations of these stratigraphic units is necessary for defining flow-field boundaries. In groundwater contamination problems, preferential pathways are controlled by the hydraulic properties of these units. This work presents and applies a Bayesian Kriging technique to a subsurface characterization problem. Expertise guesses with given uncertainties are included in the estimation procedure. This technique assures that observation data (hard data) prevail in areas close to observation points, whereas in areas without observations the guesses (soft data) have greater influence. Maps of the estimates and the associated uncertainties are shown to be key tools in reflecting the quantity and quality of the available data. The Simple and Universal Kriging become subsets of this procedure. Key words : Kriging, subsurface characterization, conditional simulation, geostatistics.