Conditional Simulation of Hydrofacies Architecture: A Transition Probability Approach

Abstract

A practical and now-popular application of geostatistics in hydrogeology is to use conditional simulation to assess impacts of subsurface heterogeneity on flow and transport processes. A transition probability approach is amenable to integration of geological concepts into analysis and modeling of spatial variability and conditional simulation of hydrofacies architecture. Hydrofacies may be defined from geological perspectives (e.g., a fluvial system with channel, splay, levee, or flood plain deposits) or textural descriptors (e.g., mud, silt, sand, gravel,…). As an alternative to variogram or covariance-based geostatistical modeling approaches, categorical geostatistical methods can be implemented with the geologically interpretable transition probability. The Markov chain model, as applied in a spatial context, is amenable to integration of geological ways of thinking as an alternative to empirical reliance on exhaustive or “training” data sets to implement geostatistical conditional simulation algorithms. Applications of conditional simulation to hydrogeological problems will benefit from using geological approaches to separate out the more deterministic and stochastic components of subsurface characterization. The overarching goal is to reduce uncertainty in flow and transport analysis and interpretation of hydrogeological data.