10 Water Resources in Karst Terranes of the Midwestern U.S.

Thursday, 2 April 2009
1:00 PM - 5:00 PM

DETECTION OF FRACTURES AND FILLED SINKHOLES WITHIN THINLY-MANTLED KARST NEAR ROCKFORD, ILLINOIS, USING GEOPHYSICS

Paper number: 10-2
Program: Symposia Sessions
Time: 1:30 PM – 1:50 PM

Authors

Abstract

Hydraulically-active fractures and sinkholes are often obscured by soil cover or vegetation, requiring the use of geophysical techniques or remote sensing for their identification. In this study electromagnetic conductivity (EM), resistivity and ground-penetrating radar (GPR) surveys were used to identify solution-enlarged bedrock fractures and filled sinkholes beneath thin unconsolidated sediment at a site approximately 10 km southeast of Rockford, IL. A ridge containing a network of mapped shallow caves, and an adjacent valley that borders a Superfund site undergoing remediation were examined. On the ridge air-filled fractures exhibit very low conductivity whereas clay-filled fractures exhibit anomalously strong attenuation of GPR signals. A trough-like feature approximately 20 m wide and 5 m deep was identified in GPR profiles obtained with both 50 and 100 MHz antennas, approximately 15 m west of the mapped cave passages. This feature may represent an elongate filled sinkhole, similar to a filled sinkhole observed along the walls of the Winquist Quarry, about 1 km to the west. Three additional 100 MHz GPR lines collected during the summer of 2008 show dipping reflections, zones of severe attenuation, and multiple diffractions, along with possible reverse-time-branches.

Within the valley glacial sediments thicken from about 1 m (near the ACME Superfund site fence) to 5-6 m near the creek, over a distance of about 300 meters. Glacial sediments are thickest near the creek. GPR profiles could not be used in the valley due to multiple reflections from overhead power-lines. However, EM and resistivity profiling provided information for fracture identification beneath 5-6 m of unconsolidated sediment (probably glacial outwash). EM profiles across the valley with a 10 m coil spacing generally fluctuated between 3 and 6 mS/m in the horizontal and vertical-dipole orientations. Modeling suggests the EM profiles were responding to the upper portions of the bedrock. Inverted Wenner-array resistivity soundings extending south from the edge of the Superfund site showed general agreement with conductivity logs obtained by a GeoProbeTM cone penetrometer, albeit with somewhat different resistivities and interface depths. The GeoProbeTM log identified bedrock at about 5.2 m depth.