Objective

There are approximately eleven million acres of land at Department of Defense (DoD) and Department of Energy sites that are highly contaminated with unexploded ordnance (UXO). The well-known high cost associated with excavating all geophysical anomalies is one of the greatest impediments to the efficient cleanup of UXO-contaminated lands. Recent live-site discrimination studies at the former Camp San Luis Obispo (SLO), CA and the former Camp Butner, NC have revealed that advanced electromagnetic induction (EMI) sensors, which currently feature multi-axis illumination of targets and tri-axial vector sensing (e.g., MetalMapper) or exploit multistatic array data acquisition (e.g., TEMTADS), together with advanced EMI models, provide superb classification performance relative to the previous generation of single-axis monostatic sensors. However, these advances have yet to improve significantly the ability to classify small targets (i.e., with calibers ranging from 20 to 60 mm) or deep targets or provide the high fidelity necessary to distinguish overlapping target signatures in highly cluttered environments. In order to achieve small- and deep-target detection and classification with 100% confidence, it is therefore necessary to take to the next level of complexity the advanced models and new-generation-sensor deployment modalities that have so far proved successful.

The objective of this project is to improve the detection and classification of small and deep targets by investigating and developing fast, noise-tolerant EMI data pre-processing and inversion approaches and extending the detection range and spatial resolution of next-generation EMI systems by using different combinations of transmitter coils adjustable in both direction and current amplitude.