High Frequency Electromagnetic Impedance Measurements For Characterization, Monitoring And Verification Efforts

Electromagnetic methods in exploration geophysics include many technologies capable of imaging the subsurface. The electromagnetic geophysical spectrum for shallow subsurface imaging is roughly 1 Hz to 500 MHz, with electrical resistivity and other geometric sounding methods located at the low frequency end and the familiar GPR method at the high end of the spectrum. Baseline studies (Pellerin et al., 1997) show that electromagnetic instrumentation in the mid- and low-frequencies (< 300 kHz) and GPR systems (> 30 MHz) are well developed in the commercial sector. In the high-frequency range of 300 kHz to 100 MHz developments have been quite recent and reside within the research community. Accurate theoretical numerical modeling algorithms are available for simulations and interpretation across the entire spectrum (Mackie and Madden, 1993; Pellerin et al., 1995; Pellerin et al., 1997; Alumbaugh and Newman, 1995; Lee et al., 1995, Newmann and Alumbaugh, 1997; Newmann, 1999; Sasaki, 1999, etc.), but instrumentation suitable for collecting calibrated field data in the important high-frequency range is critically lacking. Several attempts to develop reliable, accurate and calibrated instruments (Sternberg and Poulton, 1996; Stewart et al., 1994; Wright et el., 1996) have produced mixed results. We proposed to exploit the concept of electromagnetic impedance, the ratio of orthogonal horizontal electric to horizontal magnetic fields, to provide the necessary technology in the high-frequency band described above. The effective depth of investigation for surface impedance measurements depends on the frequency, and is commonly expressed in terms of the skin depth, the distance into the conductive half space at which the amplitude of the incoming wave has decreased to e-1 of its surface value. In order to achieve skin depths between 0.5 and 10 meters in material of resistivity between 1 and 100 ohm-m and relative permittivity between 1 and 30, frequencies bet ween about 300 kHz and 100 MHz are required. To achieve better resolution in permittivity, we need to utilize data at frequencies at the high-end of this spectrum. It is also generally true that resolution in permittivity can best be realized in an area of high electrical resistivity. To emphasize the utility and the importance of high-frequency EM measurements for mapping subsurface distribution of moisture content, we take as example a clay cap model. Figure 1 presents a simple clay cap application, based on conditions at the Savannah River Site H-Area Basin, where the specific problem is the shrinkage-induced cracks in the clay cap as it dries. Measured resistivities are about 500 W-m (Persoff, et al., 1996). The relative permittivity calculated from an experimentally validated standard mixing law (Knoll, 1996) for a clay-air-water mixture ranges from 17 or greater for a healthy clay to 12 and less for one which is too dry for regulatory compliance. Thicknesses and expected permittivities are shown for all layers of the cap. The surface ''plane-wave'' impedance, in ohms, is plotted against frequency. As is apparent, the good and bad clay conditions manifest themselves in a 27% difference in amplitude at about 30 MHz. Adjusting only the layer resistivities within the ranges observed at the site however, perturbs these curves by less than 5%. The same geometry with the observed resistivities and equal permittivities yields a nearly featureless spectrum. The change in impedance is caused solely by permittivity contrast, which cannot be detected by any conductivity-sensitive method. Variation of the clay water content yields little change in resistivity, since surface conduction dominates over saturation effects. A ground penetrating radar (GPR) survey did not 6 result in clear images in tests done at a nearby site composed of the same clays and soils. Clearly, quantitative knowledge of the layer permittivity is required to solve the problem