Paper No. 7
Presentation Time: 10:00 AM
COMPARISON OF 2D CONDITIONAL AND UNCONDITIONAL FRACTAL SIMULATIONS OF HIGHLY HETEROGENEOUS HYDRAULIC CONDUCTIVITY FIELDS
DOGAN, Mine1, MEERSCHAERT, Mark M.
2, BENSON, David A.
3, HYNDMAN, David W.
4, VAN DAM, Remke, L.
1, BUTLER Jr, James J.
5 and BOHLING, Geoffrey C.
6, (1)Department of Geological Sciences, Michigan State University, 206 Natural Science Building, East Lansing, MI 48824, (2)Department of Statistics and Probability, Michigan State University, A416 Wells Hall, East Lansing, MI 48824, (3)Hydrologic Science and Engineering, Colorado School of Mines, 1500 Illinois St, Golden, CO 80401, (4)Department of Geological Sciences, Michigan State University, East Lansing, MI 48824, (5)Kansas Geological Survey, Lawrence, KS 66047, (6)Kansas Geological Survey, University of Kansas, 1930 Constant Ave, Lawrence, KS 66047, minedoga@msu.edu
Stochastic and geostatistical methods are widely used to populate parameter fields for flow and transport models. At sites with low levels of heterogeneity in hydraulic conductivity (K), conventional approaches based on a limited number of measurements can be successful. In highly heterogeneous environments, however, novel approaches (e.g., extensive use of measured values, exploiting additional data sets, and different statistical measures) are needed to simulate parameter fields. In recent years, several novel methods have been proposed, but adequate high-resolution field data have not been available to fully test the accuracy of the simulated fields.
This study presents a comparative analysis of unconditional and conditional fractal simulations for developing realistic parameter fields along a transect through the highly-heterogeneous Macro Dispersion Experiment (MADE) site in Mississippi. The necessary high-resolution K data for this comparative analysis were obtained using a novel direct-push tool. Ground penetrating radar (GPR) facies were used as an additional soft conditioning parameter. Four K fields were simulated and compared: unconditional and conditional fractal models of the full transect, and two fields where each GPR facies has its own distinct statistical properties. These K fields were then used to parameterize a flow and transport model to analyze the effect of different simulation techniques on the concentration histories. The results were compared with the general characteristics of previous field experiments at the MADE site.
