Search

The in-phase response collected by portable loop–loop electromagnetic induction (EMI) sensors operating at low and moderate induction numbers (≤1) is typically used for sensing the magnetic permeability (or susceptibility) of the subsurface. This is due to the fact that the in-phase response contains a small induction fraction and a preponderant induced magnetization fraction. The magnetization fraction follows the magneto-static equations similarly to the magnetic method but with an active magnetic source. The use of an active source offers the possibility to collect data with several loop–loop configurations, which illuminate the subsurface with different sensitivity patterns. Such multiconfiguration soundings thereby allows the imaging of subsurface magnetic permeability/susceptibility variations through an inversion procedure. This method is not affected by the remnant magnetization and theoretically overcomes the classical depth ambiguity generally encountered with passive geomagnetic data. To invert multiconfiguration in-phase data sets, we propose a novel methodology based on a full-grid 3-D multichannel deconvolution (MCD) procedure. This method allows us to invert large data sets (e.g. consisting of more than a hundred thousand of data points) for a dense voxel-based 3-D model of magnetic susceptibility subject to smoothness constraints. In this study, we first present and discuss synthetic examples of our imaging procedure, which aim at simulating realistic conditions. Finally, we demonstrate the applicability of our method to field data collected across an archaeological site in Auvergne (France) to image the foundations of a Gallo-Roman villa built with basalt rock material. Our synthetic and field data examples demonstrate the potential of the proposed inversion procedure offering new and complementary ways to interpret data sets collected with modern EMI instruments. [ABSTRACT FROM AUTHOR]

The in-phase response collected by portable loop-loop electromagnetic induction (EMI)sensors operating at low and moderate induction numbers is typically used for sensing themagnetic permeability (or susceptibility) of the subsurface because it contains a rather small"induction" fraction depending on the electrical conductivity. The magnetization fraction ofthe in-phase signal, which is sensitive to the surrounding magnetic permeability, follows themagneto-static equations similarly to the magnetic method but with an active magneticsource. The use of an active source offers the possibility to collect data with several loop-loopconfigurations, which illuminate the subsurface with different sensitivity patterns (Klose etal., 2018). Such multi-configuration soundings thereby open the door to the imaging ofsubsurface magnetic permeability/susceptibility variations by an inversion procedure.Such an imaging procedure, which focuses on the induced magnetization effect,overcomes the classical depth ambiguity generally encountered with passive geomagneticdata.To invert multi-configuration in-phase data sets, we apply the 3D full-grid multi-channeldeconvolution (MCD) procedure presented in Guillemoteau et al. (2017). This fast imagingalgorithm allows us to invert large in-phase data sets (e.g., consisting of more thana hundred thousand data points) for a dense voxel-based 3D model of magneticpermeability subject to smoothness constraints. In this study, we first present anddiscuss synthetic examples of our inversion strategy, which aim at simulating realisticconditions. Finally, we demonstrate the applicability of our method to field data collectedacross an archaeological site in Auvergne (France) to image the foundations of agallo-roman villa consisting of basaltic rock material. Our synthetic and field dataexamples demonstrate the potential of the proposed inversion procedure offeringnew and complementary ways to interpret data sets collected with modern EMIinstruments.References:Guillemoteau, J., Christensen, N. B., Jacobsen, B. H., Tronicke, J. (2017): Fast 3Dmulti-channel deconvolution of electromagnetic induction loop-loop apparent conductivitydata sets acquired at low induction numbers. Geophysics, 82(6), E357-E369.Klose, T., Guillemoteau, J., Simon, F.-X., Tronicke, J. (2018): Toward subsurface magneticpermeability imaging with electromagnetic induction sensors: Sensitivity computation andreconstruction of measured data. Geophysics, 83, E335-E345 [ABSTRACT FROM AUTHOR]