Please provide your name and email to download


First Name

Last Name


Possibilities for UXO Classification using Characteristic Modes of the Broadband Electromagnetic Induction Resistivity

Publisher –
Zonge, 1999. Presented at “New Technology Applications Conference on the Science and Technology of Unexploded Ordnance Removal and Site Remediation,” Maui, Hawaii,  8-11 Nov., 1999

Authors –
D.D. Snyder, Scott MacInnes*, Scott Urquhart*, and K.L. Zonge, Zonge Engineering and Research Organization, Tucson, Arizona

Paper – [pdf]  UXO_1999_PossibilitiesForUXOClassification-Using_EM

The measurement of the broadband induction electromagnetic response in the form of a complex function of frequency (FEM) or, alternatively, as a transient function of time (TEM) has been applied in geophysical exploration for 30 years or more. Broadband EM methods are used in exploration in two different ways: 1) to perform “soundings” wherein the objective is to map the earth conductivity as a function of depth, and 2) for “inductive prospecting” wherein the broadband response permits the detection and characterization of large highly conductive “ore bodies” at great depth. Until recently, broadband induction EM methods were not routinely applied for shallow exploration problems. The principles of the induction EM method require that as the geometric scale of the problem decreases, there must be a corresponding broadening of the frequency range of interest in FEM systems or, equivalently, shortening of the time interval of interest in TEM systems. At present only a few field instruments are available with the requisite bandwidth for effective application to sounding or prospecting in the shallow subsurface (< 30 m).

Elementary induction EM principles are also applied in metal detectors and as such they have enjoyed a long and successful history in applications such as utilities location and in the location of metallic mines. Metal detectors are typically optimized for detecting very small objects located within a few 10′s of cm from the surface. Recently, however, new induction EM instruments have been developed specifically for shallow metal detection and site characterization. These instruments have significantly increased the depth of detection for shallow-buried metallic objects and, at least in one case, they provide measurements at more than one frequency or time delay. These instruments are now widely applied for detecting and mapping shallow-buried metal objects including UXO.

The potential for using the characteristics of the broadband induction EM response measured in the proximity of buried metallic objects to discriminate target types is generally recognized. In the context of UXO detection with TEM, McNeill et. al., concluded that “. . ., given a priori knowledge of the decay characteristics of UXO that are expected in a survey area, the evidence presented in this paper suggests that it might be possible to separate out various types of UXO (a) from each other and (b) from exploded ordnance and other trash metal.” Moreover, there is ongoing research and development directed toward developing instruments and techniques for object detection and classification with broadband induction EM.

In cooperation with Earth Tech, Zonge Engineering has been investigating how a fast TEM system might be applied for UXO characterization. In this paper, we explore how broadband induction EM responses (i.e., TEM transients, or FEM spectra) can provide a basis for UXO classification. In that regard, the next section will review briefly some important characteristics of the inductive EM response of confined conducting and permeable objects. These characteristics, long recognized by exploration geophysicists, provide a basis for expanding an FEM spectrum or TEM transient as a series of characteristic modal functions whose parameters contain information about the conductivity and size characteristics of the target. One method for the decomposition of the EM response into these characteristic modes, Prony’s method, is discussed briefly and is applied to both synthetic and real TEM transients. Finally, we present data acquired with a prototype antenna system consisting of a horizontal transmitting antenna and a 3-axis receiving antenna. With this antenna system, data were acquired using a Zonge 3-channel NanoTEM system.