Dealing Explicitly with Complexity: Evaluation of Hydrogeologic Framework Models in Capturing Subsurface Flow and Solute Transport in An Experimental Stratigraphy

Hydrogeologic framework models are routinely used to evaluate a variety of large scale geological and environmental processes impacted by groundwater flow. Effective parameters (equivalent hydraulic conductivity, macro-dispersivity) are implicit in the model construct to represent bulk flow and transport behaviors arising out of the smaller scale, unresolved heterogeneity. These parameters are of fundamental importance in hydrogeology, however, their estimation and application in modeling irregular and geologically realistic deposits remain elusive. In this study, a high-resolution, fully heterogeneous model is created by scaling up an experimental stratigraphy. It exhibits complex hydraulic conductivity (K) variations within a multi-scale framework. By conducting numerical flow experiments in this model, a novel upscaling method is developed to estimate an equivalent conductivity for each irregular shaped model unit. Two framework models are created, one of coarser stratigraphic division. By conducting flow and transport experiments in all models, the impact of effective parameterization on flow and transport predictions is elucidated. In flow prediction, the ability of the framework model to capture the bulk flow characteristics (head, streamline, flux) depends on the level of stratigraphic division, within-unit K trend, and boundary condition. In transport prediction, the first-order macrodispersion theory provides parameters that improve solute breakthrough arrival and tailing, although the shapes of the plume and the breakthrough curve are sensitive to dispersivity assignment and within-unit heterogeneity. Recently, a multi-million-cell three-dimensional stratigraphic model is created. To explore the fundamental parameterization issues in 3D, parallel flow and transport codes are currently under development.