f NOISE SOURCES AND PARAMETER ESTIMATION VARIATION IN RECENTLY DEVELOPED EMI SENSORS USED FOR UX0 DISCRIMINATION

A variety of electromagnetic induction (EMI) sensors have been recently developed for the purpose of discriminating buried unexploded ordnance (UX0) from other metallic clutter. The ultimate goal is to reduce the cost of cleaning up military ranges by positively identifying the EMI anomalies that come from dangerous UXO. This allows the clean-up operation to ignore the far larger number of anomalies that originate from other metallic clutter. These sensors are designed to collect data locally over the EMI anomaly. The data is used to identify UXO by using a physics-based model inversion. For the UXO problem, an induced dipole moment model is used and the inverted magnetic polarization factors are used to identify UXO. Errors in the inverted polarization parameters are driven by system noise and limit the ability to distinguish UXO and clutter.