Mesoscale conductivity in argillic layers at SRS

Abstract

Observations from a grid of shallow, maximum-rise piezometers at the Savannah River Site, SC indicated that ground water perching on the argillic layer was common. However, flow measurements from an interflow-interception trench indicated that lateral flow was rare and most soil water percolated through the argillic layer. We hypothesized that the lack of frequent lateral flow was primarily due to penetration of pine tree roots through the argillic layer.

Ground penetrating radar (GPR) was used to map soil structure and anomalies, such as root holes, down to two meters depth at three 10x10-m plots. These maps were used to help place 1x10-m back-hoe excavated trenches and 1x10-m auger hole transects at each of the three plots. Depth from the surface to the argillic layer was measured using a tile probe, a knocking pole and then augering to the argillic layer. The results from these three techniques were compared to each other to determine statistical differences amongst the results. Compact constant-head permeameters (CCHPs) were used to measure in-situ hydraulic conductivities in the clay-loam matrix and visually apparent anomalies in the trenches and transects. Anomalies were visually investigated by excavating with a shovel. Photographs of soil wetness were taken with a multi-spectral camera.

We discovered that all anomalies found were represented on the GPR maps, but not all predicted anomalies on the GPR maps had high conductivities. We discovered that tree root holes created anomalies, but that some conductivity anomalies were visually indistinguishable from low-conductivity soil.

Depth to clay layer was determined using a tile probe, knocking pole and an auger at 2x2-m spacing on each plot. Upon comparison, we discovered that augering was more accurate and less time-consuming, but more damaging to surrounding matrix then other techniques in these soils.