Effects of wideband complex electromagnetic properties of soils on geophysical sensor performance, The

Abstract

Common near-surface geophysical methods such as time domain electromagnetic induction (TDEM) metal detectors and ground penetrating radar (GPR) suffer performance degradation as a function of site specific complex electromagnetic soil properties (permittivity, permeability and conductivity). Knowledge of these soil properties from the kHz to the GHz frequency range can be used to predict and improve sensor performance. A prototype permittivity probe was used to measure the complex permittivity and conductivity of the soil and calculate the GPR velocity and attenuation of the from the in-situ measurements. The prototype probe was capable of accurately predicting the GPR velocities when compared with the GPR measurement and could easily predict the attenuation which is difficult to determine from actual GPR data. Unfortunately the prototype probe here has one primarily deficiency which is the assumption that the soils where it is used are non-magnetic. To illustrate the problems with using this probe in magnetic soils I made soil analogues from commercially available magnetite and crushed silica powder then measured them using a common open ended coaxial probe followed by measurements with coaxial air- line fixture which can also calculate magnetic properties. The calculated permittivities are up to twice as high when measured with the coaxial probe as they are when measured with a coaxial airline fixture which will lead to incorrect estimates of GPR velocity and attenuation. To address the performance issues of metal detectors in magnetically viscous soils I created a magnetically viscous soil analogue that could be used in mine detection training lanes instead of importing soil from sites exhibiting magnetic viscosity. Five commercially available iron oxide nano-powders were tested as additives to create the soil analogues by measuring the magnetic viscosity of these iron oxides with a new prototype instrument and compared them to samples of magnetically viscous soils collected at sites around the world. Three of the iron oxides exhibited comparable magnetic viscosities to the naturally occurring soil samples. One was selected to make a soil analogue by mixing it with crushed silica. The resulting magnetic susceptibilities compared favorably with those of the natural soil samples.