Objective

Electromagnetic (EM) signals used to locate and possibly identify unexploded ordnance (UXO) are known to propagate through earth media in linear fashion, which implies that the Green's function for EM waves can give an accurate mathematical representation of this process. Time-reversal (TR) processing of such data is based on this fact. Using large arrays of EM sources and receivers to collect a sufficiently large set of multistatic array data has shown promise for detecting and subsequently identifying the physical nature of the scattering objects. All physical objects will have potentially three dielectric values (x,y,z) associated with them, while conducting (i.e., metallic) objects will have up to an additional three (again oriented x,y,z in space) conductivity values that can be detected. TR data processing is a useful way of processing such array data and then determining physical characteristics of the detected scattering objects.

The objective of this project is to determine whether TR processing of EM data can be used successfully to improve imaging, identification, and enumeration of underground metallic objects. In particular, the TR operator (which can itself be reconstructed from EM TR data) can be used to specify the number, and also to some extent the types, of objects present and the causes of the EM scattering. Thus, the TR approach to UXO identification could be used to enumerate these objects while also determining their types (i.e., pure dielectric or metallic, UXO or non-UXO metallic objects).