The objective of this project was to demonstrate that a simple, commercially available handheld sensor (Geonics EM61-HH metal detector) could be used to discriminate between buried unexploded ordnance (UXO) and clutter. The approach used for UXO/clutter discrimination with the Multi-Sensor Towed Array Detection System EM61 array in previous demonstrations was adapted for use with the EM61-HH sensor.
The discrimination approach uses a model-based estimation procedure to determine whether or not an unknown target is likely to be an UXO. This entails estimating the size and shape of the target from the spatial pattern of the induced field above the target. The signal is a linear function of the flux through the receiving coil. If a simple guide or template is used, the position of the EM61-HH sensor as it is swept over a target can be determined with sufficient accuracy to support discrimination. This concept depends on the reliability of the sweep pattern and the existence of a fiducial mark to register the data. The pattern is fairly repeatable, and with a fiducial mark at the center, the sensor output can be matched with position in the trajectory. Sweep motion is registered by placing a small wire loop at the center of the pattern and providing a template to guide the sweep. The EM61-HH sensor measures the induced field over two time gates.
The demonstration of the EM61-HH was conducted on the Blind Grid at the Aberdeen Test Center (ATC) Standardized UXO Technology Demonstration Site. Results of controlled test-stand measurements had indicated that 95% of detected UXO could be correctly identified as UXO, while 74% of the clutter items could be correctly classified as not UXO. In the ATC field demonstration, the corresponding performance levels were reduced to 83% and 49% respectively. The average cost per target was $31.38.
Subsequent to this demonstration, the handheld sensor was used to support a UXO cleanup action on 18 acres at a site in Bridgeport, Connecticut. The net cost for doing cued identification was $115 per target. Of the 1,546 anomalies identified in the EM61 survey, 694 items were identified as being high probability of ordnance and each item was re-acquired for cued identification using the EM61-HH. Of these items, 584 were classified as high confidence clutter, 43 as uncertain, and 67 as possible UXO.
A similar process was used at the Jeep/Demolition Range at the Former Lowry Bombing and Gunnery Range in Aurora, Colorado, to identify chemical agent shipping containers. The site covered 175 acres, and 28,014 anomalies were identified in the original geophysical survey. After the initial screening, cued identification using total field magnetometer and EM61 grids was performed for over 2,000 anomalies. Based on the results of the discrimination analysis on the cued grid data, approximately 200 targets were identified as possible chemical agent shipping containers and excavations were performed. (Project Completed - 2005)