Maxwell-wagner-sillars, adsorbed water, and free-water dielectric relaxations within a hydrated arid-zone calcic soil

Abstract

Abstract-We measured the broadband (40 Hz-20 GHz) rel ative dielectric permittivity of sieved surficial samples of an arid-zone soil. The soil samples were equilibrated at 25°C with relative humidities of 43, 54, 84, and 97 %. Debye-like Maxwell-Wagner-Sillars (MWS) polarization relaxation times of 309, 227, and 73 ns were calculated for the soil samples equili brated at relative humidities of 43, 54, and 84 %, respectively.Due to electrode polarization, MWS polarization relaxations could be not be discerned on the soil sample equilibrated at 97 % relative humidity. The measured differences between the static and high-frequency relative permittivity (Es -Eoo ) due to MWS polarization were between 25 and 45 times that of the oven dry soil. This would cause the MWS relaxation to increase radar-wave attenuation rates at frequencies as high as 1 GHz.We found Debye-type relaxation times associated with adsorbed water of 13, 15, 17, 19, and 22 ps for an oven-dry sample and ones equilibrated at at relative humidities of 43, 54, 84, and 97 %, respectively. These frequencies are higher than previously thought. This progression with water contents may be due to the increase in soil solution salinity with increasing water contents.They were not likely caused by free water since a weakly resolved relaxation could be discerned corresponding to a relaxation time of approximately 8 ps, as expected for free-water.Index Terms-Adsorbed water, dielectric permittivity, free water, Maxwell-Wagner-Sillars polarization, dielectric relaxation, soil. I. I NTRODUCTIONIn earlier work based on time-domain reflectometry (TDR) measurements, we concluded that the anomalously high GPR attenuation rates found in some arid-zone soils were caused by Maxwell-Wagner-Sillars (MWS) polarization [2], [3]. Our ability to quantify the complex dielectric behavior of arid zone soils suggestive of MWS polarization had been limited. This was because some of these TDR measurements were subject to electrode-polarization effects at frequencies below I MHz. Because the apparent relaxation frequency of MWS polarization was in this frequency range, we were compelled to estimate relaxation parameters with modeling results, which may have introduced extrapolation error in these estimates.We report here broadband dielectric measurements made at low frequencies with a dielectric cell fitted to an impedance analyzer and at high frequencies with an open-ended coaxial probe fitted to a vector network analyzer. This approach extended the bandwidth and improved the precision over what could be achieved with TDR measurements. This composite approach should also provide more accurate estimates of attenuation rate over the entire GPR bandwidth. For most measurements it also shifted the onset frequency of electrode polarization far below that of the MWS, which allowed us to estimate relaxation parameters by non-linear regression fits to a simple Debye-type expression. The complex relative dielectric permittivity of a soil can

Broadband TDR permittivity spectra of lossy soils at medium to high water contents: Separation of electrode polarization from Maxwell-Wagner relaxation by modeling

“…reaching 14.33 MHz at 15.4 % for the quartz-calcite soil and 35.52 MHz at 17.0% for the quartz gypsum soil. This trend is consistent with general theory of MWR, such as formulated by Matzler [19 ] for a spherical conductive inclusion whereby ire1 is expressed as (10) in which 0"2 is the real conductivity of an inclusion, &1 ' is the real part of &1 * of the host medium, and &2 ' is that of the inclusion. Equation (1) differs little fromfrel = 1/2nT, where the relaxation time T = CT!…”

“…We used a least squares matching approach to minimize the errors for .0' and .0". Our companion paper [10] tests of one of these soils at low water contents and conductivities, and with less EP, by combining open-ended coaxial and low frequency impedance methods. The results greatly improve our previous results [3] over a far wider frequency range, and require no modeling to identify the relaxation frequencies.…”

Verification on the mode of moisture transfer in the vadose zone of thick loess

“…Gas and water were dispersed in the skeleton of unsaturated soil. Part of the water adsorbed by the soil skeleton equated to the residual water content, whereas and remaining part existed in the form of flowing water (Grant et al, 2014). The monitoring results showed that soil volumetric water content always exceeded the residual water content.…”