The main objective of the proposed work is to develop a better understanding of the behavior of diffusive electromagnetic induction (EMI) fields in underwater (UW) environments by researching EMI forward and inverse models which account for the air-water-seabed environment. This understanding will enable EMI instruments to more effectively operate UW and will enable more accurate interpretation of EMI data acquired UW. If the results of the initial 5 objectives do not indicate the targeted level of UW detection and classification for current EMI instruments, then the researchers will also investigate the optimization of UW EMI transmitter waveforms. In particular, this project proposes to:

  • Develop forward and inverse EMI models using relevant terms in the Hertzian magnetic dipole (HMD) expression to accurately account for the underlying physics of EMI fields in conducting UW environments.
  • Research the behavior of diffusive EMI fields in the air-water-seabed environment using 3D EMI solvers based on the method of auxiliary sources (MAS) and Finite Difference Time Doman (FDTD) methods.
  • Assess the UW target detection and discrimination capabilities (and limitations) of current advanced geophysical EMI sensors in the air-water-seabed environment using the FDTD method with a correspondence principle for wave and diffusion fields.
  • Account for the effects of conducting media on both the primary and secondary EMI fields and the resulting data by combining a straightforward complex image method with existing advanced forward and inverse models.
  • Compare the results of these modified methods to existing or new SERDP/ESTCP UW data sets.
  • (Contingent upon results of first 5 objectives) Research optimal transmitter current waveforms for generating a primary EM field with a maximum intensity at critical distances from the transmitter current loop to improve UW target detection and classification capabilities.

Technical Approach

This research consists of developing advanced forward and inverse EMI models for UW unexploded ordnance (UXO) detection and classification that fully account for the underlying physics of diffusive EMI field behavior in an air-waterseabed environment. EMI systems are governed by Maxwell’s equations and specifically rely on the magnetic Hertzian dipole expression for data interpretation. On land, because the conductivity of the soil is usually very low (on the order of 20mS/m), the wavenumber k in the soil is approximately zero so that contributions from the intermediate and far fields are neglected. Underwater, however, the wavenumber is much larger due to the higher conductivity of water (up to 4 S/m) and therefore the extra terms, which are proportional to k (intermediate field) and k2 (far field), in the expression for the magnetic field from a Hertzian dipole cannot be safely ignored. This leads to undulations in the field and phase perturbations which can alter data acquired from EMI instruments tailored to land based environments leading to failed data interpretations. This project will use the full theoretical expression for a magnetic Hertzian dipole, complex image theory, and both the full-3D MAS and FDTD based EMI solvers, to better understand and account for the effects of a conducting medium on both the primary and secondary EMI fields, considering the entire air-water-seabed environment.


There are approximately ten million acres of underwater areas at Department of Defense (DoD) and Department of Energy (DOE) sites that are contaminated with unexploded ordnance (UXO). It is more expensive to detect and dispose of underwater military munitions than it is to excavate the same targets on land. It is therefore beneficial to develop innovative detection and discrimination systems with minimal false alarm rates that can reliably discriminate between hazardous UXO and innocuous items. Given the advanced EMI sensors’ excellent classification performance in land-based tests, the advanced UXO detection and classification technologies have been adapted directly to underwater scenarios without considering whether to modify associated EMI models, transmitter currents wave forms, and the associated inversions schemes. As a result, recent studies for targets in a marine environment showed that extracted extrinsic and intrinsic target parameters from the data, using a standard collocated orthogonal dipole model approximation, were incorrect particularly at early time gates. To overcome these problems, this research will develop enhanced forward and inverse EMI models that account for the air water-seabed environment to enable a greater understanding of EMI data and sensors when used for UW.

  • UXO Detection and Classification,

  • Electromagnetic Induction (EMI) Properties and Sensors,

  • Electromagnetic Induction (EMI) Environmental Models,