North-Central Section - 38th Annual Meeting (April 1–2, 2004)

Paper No. 6
Presentation Time: 1:00 PM-5:00 PM

ESTIMATING STREAMBED HYDRAULIC CONDUCTIVITY BY MODEL CALIBRATION


SICKBERT, Timothy B. and PETERSON, Eric W., Department of Geography - Geology, Illinois State Univ, Campus Box 4400, Normal, IL 61761, tbsickb@ilstu.edu

The hyporheic zone—the subsurface space in which ground water and surface water mix—hosts a population of phreatobites and stygobites essential to riverine and lacustrine ecosystems. A fundamental goal of research into the hyporheic zone is to quantify the water flux between the surface and the subsurface. This flow is controlled by differences in fluid potential of the water (hydraulic head) and the ability of the material to transmit water (hydraulic conductivity). Hydraulic head is relatively easy to measure as the height of a water column; hydraulic conductivity has proven exceptionally difficult to measure in a stream bed: the literature is rich with descriptions of methods of measuring this parameter; it is almost equally rich with accounts and explanations of the difficulties. Further, most studies of saturated hydraulic conductivity focus on horizontal flow in a confined aquifer, and most studies of vertical hydraulic conductivity focus on unsaturated flow through soil. The vertical hydraulic conductivity of continuously saturated media is, therefore, less well studied.

Streams, however, naturally provide a varying head distribution which can be used to calibrate a numerical model using data collected from the stream bed material in situ. This study uses sequences of head data collected from piezometer nests in a small stream as input to a transient one-dimensional numerical finite-difference simulation. The stream incises glacial outwash comprising gravelly silty sand. The simulation was calibrated by adjusting the hydraulic conductivity to match the observed head distribution. Modeled values in the range of 10-8 to 10-6 ms-1 give the best fit of observed and calculated data for several model runs. These values are at the low end of the generally accepted range for the material, but are reasonable considering that the silt fraction controls flow and that the model calculates vertical, rather than the more generally modeled horizontal, conductivity. These results suggest that the method may be practical and useful.