Detection and classification of unexploded ordnance in challenging environments

English Abstract:

Electromagnetic induction (EMI) is the prevailing geophysical method for discriminating buried unexploded ordnance (UXO) from benign metallic clutter. EMI sensors designed for this application illuminate buried targets with primary fields from multiple directions and measure orthogonal components of the secondary field at each receiver. Subsequent processing involves fitting a physical model to observed sensor data and then using the parameters of this model to make inferences about the intrinsic physical properties of a detected target. Recent demonstration projects have shown that classification with next generation EMI sensors consistently outperforms conventional monostatic systems. 

We present applications of EMI sensing to UXO detection and classification in challenging environments. The first application is identification of very large, deeply buried (>5m) World War II era UXO in Germany. Both the depth of these ordnance and the high cultural noise encountered in urban environments precludes the acquisition of surface measurements for this problem. We describe a borehole EMI system that measures decaying secondary fields using fluxgate magnetometers deployed downhole. Transmitter loops at the surface are arranged to provide orthogonal illumination of buried targets. We show examples of field data acquired with this system and discuss the results of inversion and classification using a dipole model. 

We then consider a second challenging problem: detection and classification of small cluster munitions dropped in Laos during the Vietnam War. We present the results of a field demonstration conducted with a multi-component EMI system. The small size of the cluster munitions, magnetic soils and local topography combine to make this a challenging problem. We present processing strategies for optimizing classification performance in this scenario.