Scenario-Constrained Simplified Forward Modeling of Underwater UXO Transient Electromagnetic Secondary-Field Responses and Applicability of Symmetric-Point Differencing

This numerical study develops a scenario-specific response atlas for underwater unexploded ordnance (UXO) and evaluates the applicability of symmetric-point differencing to center-loop transient electromagnetic (TEM) profile data. The modeled configuration comprises a 5 m water layer; a highly conductive target 0.60 m long and 0.15 m in diameter; a surface center loop with a 2 m radius, 10 turns, and a 20 A current; and target-center burial depths of 0.5–2.0 m below the seabed. Target-only induced-voltage responses are calculated using a circular-loop excitation model and an orthogonal magnetic-dipole equivalent representation under the assumptions of a compact, isolated, weakly coupled target. At 0.5 ms after turn-off, inclination primarily controls peak topology, whereas azimuth controls the principal orientation of the anomaly. For this configuration, the normalized maximum response decreases from 1.00 to approximately 0.25 as burial depth increases from 0.5 to 2.0 m, yielding the empirical, model-specific coefficient k = 0.93 m−1. A separate idealized background model is used solely for a mechanistic analysis of symmetric-point differencing. Under the study-defined conservative requirements ηres ≤ 0.10 (equivalently, Rbg ≥ 90%) and Ctar ≥ 2, and with a sampling interval of Δε = 0.1, ε = 0.1 is the largest tested heterogeneity level that satisfies both the background-control and target-visibility requirements; at ε ≥ 0.3, the modeled residual becomes comparable to or larger than the target anomaly. A qualitative analysis of the idealized background over 0.5–5 ms further indicates that differencing performance must be assessed channel by channel. No tank or field observations are included. The attenuation coefficient and heterogeneity intervals are therefore theoretical outputs of this single numerical scenario rather than universal field-correction parameters.