Uxo Time-Constant Estimation From Helicopter-Borne Tem Data

Oak Ridge National Laboratory, in cooperation with Geosensors Inc., has developed a family of<br>airborne sensor systems known collectively as the Oak Ridge Airborne Geophysical System, or<br>ORAGS. The principal focus of the ORAGS research program to date has been on airborne mapping<br>and detection of unexploded ordnance, initially through magnetometry, leading to the ORAGSArrowhead<br>system and its vertical gradiometer variant ORAGS-VG. More recently the ORAGS effort<br>has been extended to include transient electromagnetic (TEM) measurements performed by the ORAGSTEM<br>system. The ongoing TEM development effort has already achieved considerable success over<br>both prepared test grids (Beard et al, in press, Beard et al 2003) and a former bombing site (Doll et al,<br>2003) at the former Badlands Bombing Range (BBR) in South Dakota. Unexploded ordnance (UXO)<br>objects ranging in size from 250 pound bombs down to 61mm mortar rounds, 60 mm illumination shells<br>and 2.75 inch rocket components were detected by both magnetometer and transient electromagnetic<br>technologies during the BBR trials conducted in September, 2002.<br>The signal/noise ratio (SNR) observed in TEM measurements during these trials was high<br>enough to motivate investigation of the target decay information contained in the measured transients.<br>An improved transient analysis technique based on the Matrix Pencil Method was developed to improve<br>the accuracy of exponential decomposition of the observed transients. Where SNR was satisfactory, this<br>method yielded repeatable results that reliably distinguished compact, long-time-constant targets such as<br>bombs and artillery shells from short-time-constant targets such as thin-walled scrap from practice<br>bombs. This step marks another milestone toward the goal of rapid, detailed mapping and discrimination<br>of unexploded ordnance based on airborne surveys. As system sensitivity and resolution continues to<br>improve, target discrimination methods are expected to become standard data analysis tools.