ABSTRACT

The unknown heterogeneity of the subsurface remains a major obstacle to reliable simulation of subsurface flow and transport processes. Characterization of spatial patterns in properties with typically sparse data requires knowledge of the geologic processes that created the patterns. A geostatistical approach based on transition probability theory provides a means for quantitative modeling of three-dimensional hydrostratigraphy through the use of commonly available field data as well as geologic fundamentals and knowledge of the depositional processes. By modeling inter-facies transition probabilities with Markov chains, an intuitive method for building three-dimensional models from basic geologic principles is developed, extending qualitiative geologic characterization into the quantitative realm necessary for flow and transport simulation. This geologic/geostatistical technique, implemented with the software TProGS, uses the hard data together with interpretive input on proportions, average lengths, and juxtapositioning of geologic facies to create multiple realizations of heterogeneity. The resulting characterizations honor fundamental probability laws while preserving observed or inferred facies proportions, continuity, asymmetries (e.g., fining upward sequences), and facies relations (e.g., levee adjacent to channel facies). Lateral facies relations, which are typically undersampled, can be modeled based on the observed vertical patterns in facies through the use of Walther's Law. Furthermore, the approach can incorporate nonstationarities such as spatially varying dip angles, or more severe nonstationarity, such as unconformities and transitions between different depositional environments. Example applications in alluvial fan and fluvial environments depict the resulting hydrostratigraphic models and demonstrate the significance of highly resolved heterogeneity in the simulation of flow and transport, while providing a fair degree of field validation with hydraulic and groundwater age-date data.   

FUTURE SEMINARS PAST SEMINARS

[ GG home | SOEST home | GG People ]