Improving the reliability of soil EC-mapping: Robust apparent electrical conductivity (rECa) estimation in ground-based frequency domain electromagnetics

The use of frequency domain electromagnetic (FDEM) instrumentation in ground-based near-surface applications is continually expanding. In soil studies, FDEM is mainly deployed to inform on soil properties such as salinity, texture, moisture or compaction by evaluating the low induction number (LIN) apparent electrical conductivity (ECa). However, advanced FDEM applications face terrains and conditions in which these LIN approximations are no longer valid. Under such conditions, the aforementioned instrumentation output systematically underestimates ECa values; making semi-qualitative data analysis, representation and practical interpretation impossible. For this reason, methods have to be developed to relate the complex FDEM responses to a reliable ECa, beyond the limitations of the LIN approximation. In this work, we present a quadrature-phase algorithm (i.e. excluding the in-phase signal from the estimation) to obtain a robust apparent electrical conductivity at both low and high induction numbers. In addition, the algorithm classifies the ECa estimations as either robust or non-robust based on instrument noise, desired precision of the ECa estimation and the effect of subsurface magnetic variations. The algorithm accounts for electrical conductivity, instrument elevation, coil geometry, operating frequency, subsurface magnetic variation and random errors. The presented procedure, which was tested on synthetic and experimental data, consists a robust and straightforward approach to increase reliability of ECa derived soil mapping.