2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 1
Presentation Time: 1:15 PM

MULTISCALE AQUIFER HETEROGENEITY – INSIGHTS FROM ‘JURASSIC TANK’


DAVIS, J. Matthew, Department of Earth Sciences, Univ of New Hampshire, Durham, 03824, matt.davis@unh.edu

The deposits generated in the Experimental EarthScape Basin at the University of Minnesota (Paola and others) provide an opportunity to study, at unprecedented resolution, the spatial patterns of sedimentary materials in three dimensions. The deposits generated from the 1999-2000 experiment consist of a bimodal mixture of coal and sand that mimics deposition of fine- and coarse-grained materials, respectively. The braidplain facies of the experimental deposits exhibit a channel-interchannel-overbank architecture similar to the architecture observed in outcrops of the Albuquerque Basin.

The grayness of the experimental deposits is interpreted as a simple mixture of the white sand and black coal. Then, using a fractional packing model, the fine and coarse fractions are mapped into a permeability field, resulting in a high-resolution numerical model of an alluvial aquifer. The aquifer model is useful both for studying the spatial statistical properties of log-k and for modeling of fluid flow and solute transport through a system with complex sedimentary structures.

The spatial statistical analysis suggests that the deposits exhibit multiple scales of heterogeneity that can be modeled with an anisotropic monofractal. The length scales that give rise to the fractal structure appear to be closely related to the sedimentary architecture of the deposits. Fluid flow and solute transport modeling illustrate the affect that these fractal structures have on hydrologic processes.

In addition to the high-resolution modeling, experimental deposits will also provide a benchmark for testing and refining various upscaling models, such as those used for permeability and dispersivity. While much insight can be gained from studying experimental deposits within their own context, applying these results to specific aquifers will require an improved understanding of the scaling relationships between the experimental and natural systems.