Research Scientist and Laboratory Director, Northeastern University, Boston, MA
Associate Professor, Northeastern University, Boston, MA
Professor, Northeastern University, Boston, MA
(Received 21 December 2004; accepted 19 September 2005)
Cross-well radar (CWR) uses electromagnetic wave antennas lowered into sampling wells, to image the dielectric properties of soil and detect scattering objects such as contaminants (in this case, dense nonaqueous phase liquids (DNAPLs)). To better understand the physics of CWR in soils, it is necessary to experimentally evaluate the behavior of antennas and electromagnetic (EM) waves. A proto-type model of an infinite soil medium was experimentally simulated by constructing a pilot-scale facility, referred to as SoilBED. The most important issue in any experimental research is the repeatability, reproducibility, and reliability of the results. This paper evaluates different factors affecting experimental data collection in order to achieve the required specifications to collect reproducible and reliable data. Antenna depth and insertion problems, soil disturbance, and boundary condition effects were experimentally evaluated. Other (desired or undesired) transmission couplings were studied, and efforts were conducted to eliminate the undesired paths. The calibrated and validated setup can be and is used for different purposes other than DNAPL detection, such as soil and antenna characterization, theoretical simulation validation, and inverse scattering by various materials in dry or saturated soil backgrounds.
Paper ID: GTJ14000