INTEGRATING TILL FRACTURES INTO NUTRIENT TRANSPORT MODELS AT PLOT AND WATERSHED SCALES: A DUAL CONTINUUM APPROACH

Contamination of surface water and groundwater from intensive agriculture is a major water-quality issue in much of the upper Midwest. Previous work has shown that till in Iowa (and elsewhere in the Midwest) is densely fractured (> 643 fractures/m³) to depths of 4 m, and that fractures are present and likely active to greater depths. Because groundwater velocities are one to three orders of magnitude higher in the fractures than in the unfractured matrix, fractures effectively control nutrient transport in till. Unfortunately, fracture flow is not routinely integrated into watershed-scale transport models, most likely due to insufficient fracture data and computational demands; i.e., modeling fracture flow in millions of fractures within a single field is computationally difficult. However, ignoring this transport mechanism could result in a model that does not accurately forecast site hydrology and water quality. A dual-continuum model, where flow and transport are simulated in both a matrix and fracture domain simultaneously, is being tested to simulate the role of fracture-enhanced nutrient transport while maintaining computational viability. Fracture orientation, intensity, and aperture data were taken from previously published field studies. Hydraulic conductivities (K) of the matrix and fracture domains of the till were computed using methods developed within the ISU Hydrogeology Research Group. A 3-D, variably-saturated, finite-element model, HydroGeoSphere, will be used to model groundwater, surface water, and climate processes in a test area – the 500-acre Walnut Creek watershed south of Ames, Iowa – where water level, climate, land use, and fertilizer application data are available for a 26-year period. Prior to those simulations, we plan to implement the model at the plot scale on a farm to determine number of model layers, appropriate boundary conditions, and the range of parameters. Soil properties will be populated from SSURGO. Bulk till and fracture K values will be taken from current and previous work. A high-resolution DEM will be used for surface topography. Our goal is to demonstrate that including fractures in watershed-scale nutrient-transport models in till environments will result in improved predictions of water quality.