Electrical spectral induced polarization (SIP) measurements are increasingly being used in environmental studies of the saturated zone. To better understand the mechanisms causing polarization and to extend the range of SIP applications to the vadose zone, it is important to investigate how the SIP response is affected by water saturation. Therefore, sand and sand‐clay mixtures were drained in several steps using a novel measurement set‐up allowing SIP measurements with a high accuracy. The measured SIP spectra were interpreted by Debye decomposition, which provided a relaxation time distribution and a chargeability distribution that was converted to a normalized total chargeability. The results showed that the normalized total chargeability of the fully saturated samples increased with increasing clay content due to the larger specific surface area of the clay minerals. Furthermore, normalized total chargeability first increased and then decreased with decreasing saturation for the pure sand and the 5% sand‐clay mixture. The normalized total chargeability values for the 10% and 20% sand‐clay mixtures only decreased with saturation. The peak relaxation time of the sand‐clay mixtures clearly decreased with decreasing saturation. Existing grain‐size based mechanistic models for SIP are not able to explain the observed behaviour and the observed relationship between relaxation time and saturation suggests that model concepts relying on polarization processes in the pore space are warranted to explain the measurements on variably saturated porous media presented here.

The concept of specific polarizability [Formula: see text], being the ratio between imaginary conductivity and specific surface area, can be used to represent the polarization of the mineral-fluid interface per unit pore-volume-normalized surface area [Formula: see text] and to account for the control of the fluid chemistry and/or mineralogy on induced polarization (IP) measurements. We used a database of IP measurements on sands and sand-clay mixtures to investigate the variation in [Formula: see text] as a function of clay content and/or mineralogy. We found an apparent variation in [Formula: see text] between sands and sand-clay mixtures when [Formula: see text] was calculated using the nitrogen adsorption (Brunauer-Emmett-Teller — BET) method, with clays having an apparently higher [Formula: see text] than sands. However, this variation was considerably reduced when [Formula: see text] was calculated using a wet-state methylene blue (MB) method that also sensed the surface area associated with internal layers of clay minerals inaccessible with the dry-state BET method. Furthermore, the imaginary conductivity was significantly better correlated with [Formula: see text] determined from the MB method relative to [Formula: see text] determined from the BET method. We found no evidence for a strong difference in the specific polarizability of quartz and clay minerals. This finding contradicted predictions from recent mechanistic formulations of the IP response of the Stern layer. Our findings have significant implications for improving and simplifying the interpretation of IP measurements in near-surface materials.

The prediction capabilities of unsaturated flow and transport models are often limited by insufficient knowledge about the structural and textural heterogeneity of the soil. To obtain more information about soil structure, texture, and heterogeneity, as well as hydraulic properties, noninvasive spectral induced polarization (SIP) measurements have shown promise. However, there clearly is a need for more detailed investigations on the relation between SIP and hydraulic properties, in particular for unsaturated soil. To address this need, we developed a novel experimental setup that allows simultaneous determination of hydraulic and electrical properties. Our setup consists of multistep outflow equipment for hydraulic measurements and an electrical impedance spectrometer for measurements of the complex electrical conductivity. Two different measurement cells were constructed. The first measurement cell can only be used for nonshrinking soil, while the other one also allows measurements on mildly shrinking soil. Test measurements showed that the measurement accuracy is very high for the complex electrical conductivity. Nonpolarizable point electrodes are used for voltage measurements, ensuring a good contact to the sample for a wide range of water saturations. An experiment performed on unconsolidated sand investigated the dependence of the complex electrical conductivity and fitted Cole–Cole parameters on water saturation. The phase values and chargeability increased with decreasing saturation. The relaxation time associated with a phase peak was independent of saturation and corresponded reasonably to the average grain size of the sand. Therefore, no relationship between relaxation time and unsaturated hydraulic conductivity was found for this well‐sorted sand.

No abstract

A: ERT, electrical resistivity tomography; RGB, red, green, blue; TDR, time domain refl ectometry.O R Temporal mapping of the three-dimensional spa al distribu on of solute transport in soils is needed for an improved understanding of the underlying processes. Numerous studies have confi rmed that Brilliant Blue imaging provides spaally highly resolved informa on on solute transport in soils. A drawback of the method, however, is its destruc ve character, which prevents three-dimensional mapping of the temporal evolu on of Brilliant Blue plumes or fronts. In this study, we determined that the nega ve ionic charge of the Brilliant Blue molecule for moderately acid and basic environments provides an electrical conduc vity contrast that can be detected by means of me domain refl ectometry (TDR) and electrical resis vity tomography (ERT). Time-lapse ERT supplies three-dimensional spa otemporally resolved image data through minimally invasive measurements but with limited spa al resolu on. A joint applica on of Brilliant Blue imaging and ( me-lapse) ERT off ers poten al benefi ts such as ERT image valida on by dye staining, spa otemporally resolved Brilliant Blue displacement studies, and improved ERT inversion regulariza on based on Brilliant Blue image-derived solute plume or front characteris cs. This study illustrates the eff orts needed to quan ta vely relate Brilliant Blue concentra on and electrical conduc vity as well as the poten al and limita ons of noninvasive smoothness-constrained ERT for solute imaging in soils.