Time domain reflectometry measurement principles and applications

SCIENTIFIC BRIEFING HYDROLOGICAL PROCESSES Hydrol. Process. 16, 141–153 (2002) DOI: 10.1002/hyp.513 Time domain reflectometry measurement principles and applications Scott B. Jones,1 * Jon M. Wraith2 and Dani Or1 1 Department of Plants, Soils and Biometeorology, Utah State University, Logan, UT 84322-4820, USA 2 Department of Land Resources and Environmental Sciences, Montana State University, Bozeman, MT 59717-3120, USA *Correspondence to: S. B. Jones, Department of Plants, Soils and Biometeorology, Utah State University, Logan, UT 84322-4820, USA. E-mail: sjones@mendel.usu.edu Abstract Time domain reflectometry (TDR) is a highly accurate and automatable method for determination of porous media water content and electrical conductivity. Water content is inferred from the dielectric permittivity of the medium, whereas electrical conductivity is inferred from TDR signal attenuation. Empirical and dielectric mixing models are used to relate water content to measured dielectric permittivity. Clay and organic matter bind substantial amounts of water, such that measured bulk dielectric constant is reduced and the relationship with total water content requires individual calibration. A variety of TDR probe configurations provide users with site- and mediaspecific options. Advances in TDR technology and in other dielectric methods offer the promise not only for less expensive and more accurate tools for electrical determination of water and solute contents, but also a host of other properties such as specific surface area, and retention properties of porous media. Copyright  2002 John Wiley & Sons, Ltd. Key Words TDR; dielectric; permittivity; soil moisture; soil electrical conductivity Introduction Time domain reflectometry (TDR) is a relatively new method for measurement of soil water content and electrical conductivity. Each of these attributes has substantial utility in studying a variety of hydrologic processes. The first application of TDR to soil water measurements was reported by Topp et al. (1980). The main advantages of TDR over other soil water content measurement methods are: (i) superior accuracy to within 1 or 2% volumetric water content; (ii) calibration requirements are minimal— in many cases soil-specific calibration is not needed; (iii) lack of radiation hazard associated with neutron probe or gamma-attenuation techniques; (iv) TDR has excellent spatial and temporal resolution; and (v) measurements are simple to obtain, and the method is capable of providing continuous measurements through automation and multiplexing. A variety of TDR systems are available for water content determination in soil and other porous media (e.g. Figure 1). Many, but not all commercially available systems may also be used to measure soil electrical conductivity. Thus potential users should consider present and future measurement requirements before purchasing. Copyright  2002 John Wiley & Sons, Ltd. 141 Received 10 June 2001 Accepted 30 July 2001 S. B. JONES, J. M. WRAITH AND D. OR Figure 1. Different TDR devices shown are: (a) 1502C (Tektronix Inc., Beaverton, OR); (b) TRIME-FM (IMKO, Ettlingen, Germany); (c) TRASE System I (Soil Moisture Equipment Corp., Goleta, CA); (d) TDR100 (Campbell Scientific Inc., Logan, UT) Basic principles In the telecommunications industry TDR is used to identify locations of discontinuities in cables, hence the term ‘cable tester’ is a common name for generalpurpose TDR instruments. The signal propagation velocity Vp (a function of the cable dielectric constant), along with a typical reflection at a point of discontinuity in a cable, allows the operator to determine locations of line breaks or other damage to cables using travel time analysis. Using similar principles, a waveguide or probe of known length L may be embedded in soil (Figure 2) and the travel time for a TDR-generated electromagnetic ramp to traverse the probe length may be determined. From the travel time analysis the soil’s bulk dielectric constant is computed (the terms dielectric constant and dielectric permittivity are synonymous), from which the volumetric water content is inferred. The bulk dielectric constant εb of soil surrounding the probe is a function of the propagation velocity (v D 2L/t) according to εb D
c 2  D  ct 2L 2 ⊲1⊳ where c is the speed of light (velocity of electromagnetic waves) in vacuum (3 ð 108 m s1 ), and t is the travel time for the pulse to traverse the length of the embedded waveguide (down and back: 2L). The travel time is evaluated based on the ‘apparent’ or electromagnetic length of the probe, which is characterized on the TDR output screen by diagnostic Copyright  2002 John Wiley & Sons, Ltd. changes in the waveform: x1 marks the entry of the signal to the probe, and x2 marks the reflection at the end of the probe (Figure 3). As shown in Figure 3, the apparent probe length (x2  x1 ) increases as the water content (and dielectric constant) increases, a consequence of reduced propagation velocity. The relationship between locations of the two reflection points x1 and x2 , and the bulk dielectric constant is:   x2  x1 2 ⊲2⊳ εb D Vp L where Vp is a user-selected relative propagation velocity, often set at Vp D 0Ð99. For most practical applications it suffices to consider the definition of the dielectric constant in Equation (1), which simply states that the dielectric constant of a porous medium is the ratio squared of the propagation velocity in vacuum relative to that in the medium. The soil’s bulk dielectric constant εb is dominated by the dielectric constant of liquid water εw D 81 (20 ° C), as the dielectric constants of other soil constituents are much smaller; e.g. soil minerals εs D 3 to 5, frozen water (ice) εi D 4, and air εa D 1. This large disparity of dielectric constants makes the method relatively insensitive to soil composition and texture, and thus a good method for ‘liquid’ water measurement in soils. Note that, because the dielectric constant of ice is much lower than for liquid water, the method may be used in combination with a neutron probe or other techniques that sense total soil water content to determine separately the volumetric liquid 142 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING TDR Cable Tester (Tektronix 1502B) BNC Connector Waveform 3-Rod Probe Figure 2. TDR cable tester with three-rod probe embedded vertically in surface soil layer and frozen water contents in frozen or partially frozen soils (Baker and Allmaras, 1990). Several factors influence dielectric constant measurements, including soil porosity and bulk density, measurement frequency, temperature, water status (bound or free) and dipole moments induced by mineral, water, and air shapes. The need to relate water content to εb and to account for the factors just mentioned has resulted in a variety of empirical and ‘dielectric mixing’ models. Empirical and dielectric mixing models Two basic approaches have been used to establish the relationships between εb and volumetric soil water content . The empirical approach simply fits mathematical expressions to measured data, unique to the physical characteristics of the soil or porous medium. Such an approach was employed by Topp et al. (1980), who fitted a third-order polynomial to the observed relationships between εb and  for multiple soils. The empirical relationship for mineral soils Copyright  2002 John Wiley & Sons, Ltd. proposed by Topp et al. (1980) of v D 5Ð3 ð 102 C 2Ð92 ð 102 εb 5Ð5 ð 104 ε2b C 4Ð3 ð 106 ε3b ⊲3⊳ provides an adequate description for the water content range <0Ð5, which covers the entire range of interest in most mineral soils, with a  estimation error of about 0Ð013. However, Equation (3) fails to describe the εb   relationship adequately for water contents exceeding 0Ð5, and for organic soils or mineral soils high in organic matter or clay content. In its simplest form the dielectric mixing approach uses dielectric constants and volume fractions for each of the soil constituents (e.g. solid, water, air) to derive a relationship describing the composite (bulk) dielectric constant. Such a physically based approach was adopted by Birchak et al. (1974), Dobson et al. (1985), Roth et al. (1990), and Friedman (1998). According to Roth et al. (1990), the bulk dielectric 143 Hydrol. Process. 16, 141–153 (2002) S. B. JONES, J. M. WRAITH AND D. OR 1.25 Reflection coefficient 1.00 εb = (x 2 − x 1) 2 L . νp θv = 0 0.75 0.1 0.50 0.2 0.25 X2 0.00 0.3 −0.25 X1 −0.50 2.50 2.75 3.00 3.25 3.50 Apparent distance [m] Figure 3. Sample TDR waveforms measured in Millville silt loam soil using the Tektronix 1502B TDR cable tester and three-rod probes (length L D 0.15 m). The calculation of the bulk dielectric constant is based on εb D [⊲x2  x1 ⊳/⊲LVp ⊳]2 , where Vp is the relative velocity of propagation (usually set at 0.99) constant of a three-phase system may be expressed as 1/ˇ  εb D εˇw C ⊲1  n⊳εˇs C ⊲n  ⊳εˇa ⊲4⊳ where n is the soil’s porosity, 1 < ˇ < 1 summarizes the geometry of the medium in relation to the axial direction of the wave guide (ˇ D 1 for an electric field parallel to soil layering, ˇ D 1 for a perpendicular electrical field, and ˇ D 0Ð5 for an isotropic two-phase mixed medium); 1  n,  and n   are the volume fractions and εs , εw and εa are the dielectric constants of the solid, water and air phases respectively. Note that the summed volumetric contribution of each phase relative to the total volume is unity. Rearranging Equation (4) and solving for  yields ˇ ε  ⊲1  n⊳εˇs  nεˇa ⊲5⊳ D b ˇ ˇ εw  εa which determines the relationship between  and εb measured by TDR. Many have used ˇ D 0Ð5, which was shown by Roth et al. (1990) to produce a calibration curve very similar to the third-order Copyright  2002 John Wiley & Sons, Ltd. polynomial proposed by Topp et al. (1980) for the water content range 0 <  < 0Ð5. If we introduce into Equation (5) common values for the various constituents, such as ˇ D 0Ð5, εw D 81, εs D 4, and εa D 1, we obtain an approximate simplified form: D p εb  ⊲2  n⊳ 8 ⊲6⊳ Note that the soil’s porosity must be known or estimated when using the mixing model approach. A comparison between the Topp et al. (1980) expression [Equation (3)] and a calibration curve based on Equation (5) with n D 0Ð5 and ˇ D 0Ð5 is depicted in Figure 4. Soils having high clay or organic matter content often require soil-specific calibration. The presence of high porosity combined with large amounts of ‘bound’ water fraction produces substantial differences between the dielectric signature of typical mineral versus organic soils. This is illustrated (Figure 4) by comparing an empirical expression of Schaap et al. (1996) fitted to 505 TDR measurements in organic 144 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING 80 70 Bulk dielectric constant Topp et al. 60 Practical range of water content measurement in mineral soils 50 40 30 Mixing model 20 Forest soil organic horizons 10 0 0.0 0.1 0.2 0.3 0.4 0.5 Volumetric water content 0.6 [cm3 0.7 0.8 0.9 cm−3] Figure 4. Relationships between bulk soil dielectric constant and  expressed using the Topp et al. (1980) empirical equation for mineral soils and for the Birchak et al. (1974) mixing model using a porosity of 0Ð5. Higher porosities and bound water in forest soil organic horizons exhibit a reduced dielectric signature by comparison (Schaap et al., 1996) horizons of eight forest floor soils given as p  D ⊲0Ð133 εb  0Ð146⊳0Ð885 ⊲7⊳ Media with large porosities, such as horticultural planting media (n ³ 0Ð95) and clays, are susceptible to large variations in bulk density and porosity. Independent measurements of bulk density should improve calibration, but the bound water influences, which are generally correlated to surface area, may be more difficult to account for. Bound water and temperature effects Water molecules within the vicinity of solid surfaces are subjected to interfacial forces that constrain their movement, rendering them rotationally hindered. Consequently, both their ability to follow the alternating electric field, and their dielectric constant are reduced. High surface-area porous media (e.g. clays, peats, some forest soils) bind a substantial fraction of the water phase, resulting in reduced bulk dielectric constant measurements relative to low surfacearea materials (e.g. sandy soils) when compared at Copyright  2002 John Wiley & Sons, Ltd. similar water contents. The amount of bound water is related to the specific surface area and the thickness of the affected (bound) layer. Estimates of the dielectric constant values for the first, second and third monolayers of water are available (Thorp, 1959; Bockris et al., 1963). Or and Wraith (1999) developed a model for the temperature-dependent thickness of the bound water phase. For media with large surface areas it was found that TDR-measured water content increased with increasing temperature due to reduction of the bound water layer thickness, without any changes in the total mass of water (an example is depicted in Figure 5). This phenomenon has important implications for interpretation of near-soilsurface measurements using TDR (e.g. infiltration or evaporation studies), where the measurement artefact must be filtered out. Interestingly, measurements of the dielectric response of wetted porous media under changing temperature may be useful in estimating the specific surface area (m2 kg1 ), because of the fundamental relationship between bound water and solid surface area (Wraith and Or, 1998). Expressions have been derived for describing the combined bound- plus 145 Hydrol. Process. 16, 141–153 (2002) S. B. JONES, J. M. WRAITH AND D. OR 0.200 Water content [m3 m−3] Water content Electrical conductivity 0.175 0.4 0.150 0.3 0.125 0.2 0.100 55 45 35 25 15 5 Electrical conductivity [dS m−1] 0.5 0.1 Soil temperature °C Figure 5. Measured volume water content and bulk soil electrical conductivity within sealed soil containers under imposed temperature changes. Soil temperatures are indicated; actual water content did not change. Modified from Wraith and Or (1999), copyright 1999 by the American Geophysical Union free-water dielectric constant based on surface area and bulk density (Friedman, 1998) and including the temperature dependence (Jones and Or, 2001). TDR probe configurations Modelled after a coaxial design, Campbell (1990) and Heimovaara (1994) used seven-wire probes to measure the dielectric constant of soils and liquids. A number of different geometrical configurations have been proposed, which have a single central conductor and from one to six outer conducting rods (Figure 6). The two-wire probe has the advantage of minimal soil disturbance, but produces an unbalanced signal, leading to unwanted noise and signal loss (White and Zegelin, 1995). This problem may be minimized using a balun (balancing transformer) embedded in the probe head to reduce signal and information loss (Spaans and Baker, 1993). The three- or more-rod probes provide a balanced signal, avoiding the balun requirement at the expense of some additional soil disturbance [see Zegelin et al. (1989) for a comparison]. Though not commonly used in soils, the parallel plate probe was shown by Robinson and Friedman (2000) to provide a highly uniform electrical field between plates. The highly concentrated electrical field converging on the central conductor Copyright  2002 John Wiley & Sons, Ltd. of the multi-wire probes (see Figure 6), more heavily weights the dielectric constant of constituents within this region. Ferre et al. (1998) found that two-rod probes have a larger sample area compared with threerod probes, and that thin rod coatings (for reducing conductive losses) for any probes will reduce the sampling area of the probe. Measurement error increases, for example, as air gaps develop when probes are repeatedly inserted and removed, or in shrink–swell soils. For all conventional probe designs, water content is often assumed to be uniformly integrated along the probe’s longitudinal axis. A recent study by Chan and Knight (1999) cautions against the accepted notion that, if water is evenly distributed along the rods or concentrated in one or more ‘pockets’, the same measured mean dielectric constant results. Their analysis is based on the wavelength  to layer (heterogeneity) thickness t ratio /t. For /t > 10 the dielectric constant of the soil is computed as an arithmetic average (effective medium theory) of the layers, whereas for /t < 1 the geometric average (Ray theory) of the soil layers is used to compute the soil dielectric constant. Scattering effects that occur within the transition zone, 1 < /t < 10, may cause measurement difficulties, and the propagation direction of the 146 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING Figure 6. Various TDR soil moisture probe designs have been proposed, such as the multi-wire and parallel plate configurations shown here. Electrical field lines generated for different probe configurations are also shown, where closer line spacing is associated with a more concentrated field (i.e. greater influence on permittivity) electromagnetic wave relative to the layering is also an important factor. The particular spatial sensitivities of different probe configurations can be used to one’s advantage in specific research applications. For example, a two- or three-rod probe placed horizontally serves as an effective point (plane) measurement for water or solute fronts moving vertically through soil profiles. Sevenrod or parallel plate designs, on the other hand, sample a larger effective volume of soil, which may be desired for routine measurements. Other TDR probe applications have been developed that provide soil matric potential measurements (e.g. Wraith and Or, 2001). These may be used separately, or may be paired with conventional probe designs to obtain simultaneous in situ measurements of  and matric potential h from which the soil water characteristic relationship ⊲h⊳ may be elucidated. Copyright  2002 John Wiley & Sons, Ltd. TDR measurements are easily automated using a computer or datalogger, and analysis of waveforms is commonly completed during measurement, or may be analysed later if waveforms are saved. Automated measurement of  requires only a few seconds for each probe. As many as 8 or even 16 probes may be attached to a single multiplexer, and several multiplexers may be connected in series to provide a large array of spatially distributed measurements. However, owing to signal deterioration with cable length, practical distances from the probe to TDR unit are typically limited to 20 to 30 m. Longer cable lengths may provide reliable readings where soil salinity and clay content are low. The ability to obtain highresolution time series measurements at multiple locations (e.g. depths) using automated and multiplexed TDR is a particularly useful research and management tool. 147 Hydrol. Process. 16, 141–153 (2002) S. B. JONES, J. M. WRAITH AND D. OR Measurement of salinity and ionic solutes Dalton et al. (1984) also first demonstrated the utility of TDR to measure the apparent bulk soil electrical conductivity. This unique ability to measure both soil water content  and apparent soil electrical conductivity ECa using the same instrumentation and probes, and in the same soil volumes, provided new opportunities to investigate salinity and the behaviour of ionic solutes in soils. The most critical frequencies for measurement of dielectric constant based on TDR travel-time are near 1 GHz, whereas TDR ECa measurement relies on the lowest frequencies available (low kilohertz range), as close to direct current (DC) as feasible. Measurement of electrical conductivity using TDR is based on attenuation of the applied signal voltage as it traverses the medium of interest. As the transverse electromagnetic waves propagate along TDR probes buried in soil, the signal energy is attenuated in proportion to the electrical conductivity along the travel path. This proportional reduction in signal voltage is accurately related to the bulk soil electrical conductivity. Comparisons between the electrical conductivity measured in solutions using both TDR and standard methods have repeatedly demonstrated the potential accuracy and precision of TDR measurements (e.g. Spaans and Baker, 1993; Heimovaara et al., 1995; Mallants et al., 1996; Reece, 1998). TDR electrical conductivity measurements may also be easily automated and multiplexed in the same manner as for . However, some commercially available TDR equipment does not include provision to measure electrical conductivity. Originally proposed by Giese and Tiemann (1975), the thin-section approach has been shown to be a particularly effective means of quantifying ECa using TDR. The Giese and Tiemann equation may be written as:   ε0 c Z0 2V0 1 ⊲8⊳ EC⊲S m1 ⊳ D L Zc Vf where ε0 is the dielectric permittivity of free space (8Ð9 ð 1012 F m1 ), c is the speed of light in vacuum (3 ð 108 m s1 ), L (m) is probe length, Z0 () is the characteristic probe impedance, Zc is the TDR cable tester output impedance (typically 50 ), V0 is the incident pulse voltage and Vf is the return pulse voltage after multiple reflections have died out 0.4 ECW = 0 dS/m Reflection coefficient 0.2 0.0 −0.2 3 dS/m −0.4 6 dS/m −0.6 V0 −0.8 12 dS/m V1 Vf −1.0 0 10 20 30 40 50 Travel time [ns] Figure 7. Attenuation of the TDR signal due to increasing soil solution electrical conductivity results in a reduced reflection coefficient at large distances (i.e. low frequency). The end-of-probe reflection is indistinguishable above about 6 dS m1 , where time domain analysis for soil water content determination fails Copyright  2002 John Wiley & Sons, Ltd. 148 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING (Figure 7). The quantity ε0 c/L in Equation (8) may be simplified to 1/(120L), and the values of V0 and Vf are easily acquired from the TDR output signal. A separate calibration procedure is required to determine the probe characteristic impedance Z0 . This may be determined by immersion of the probe in deionized water with known (e.g. Weast, 1986) dielectric constant ε according to:   p V1 Z0 D Zc ε ⊲9⊳ 2V0  V1 where the required signal voltages V0 and V1 are illustrated in Figure 7. For specific probe geometries (e.g. true coaxial, two-rod balanced design), Z0 may be calculated with formulae found in electronics textbooks, but because of inconsistencies in probe manufacture we prefer to use the simple deionized water calibration procedure. Finally, ε0 cZ0 /L D Z0 /120L may also be lumped into a geometric probe constant K, and EC may thereby be estimated using Equation (8) as   K 2V0 EC D ⊲10⊳ 1 Zc Vf with K either calculated using Equation (8) and the relevant physical quantities (ε0 , c, L), or empirically determined by immersing the probe in one or (preferably) more solutions of known EC, and using K D ECref ZL fT ⊲11⊳ with ECref the known electrical conductivity of the reference solution, ZL as the measured resistive load impedance across the probe [ZL D Zc /⊲2V0 /Vf  1⊳] and fT a temperature correction coefficient to relate the measured reference solution to a desired standard temperature. Heimovaara et al. (1995) found that the relationship fT D 1/[1 C 0.019⊲T  25⊳] was appropriate for a variety of saline solutions, using 25 ° C as the standard temperature. The unique ability of TDR to measure both  and EC provides many opportunities for research. For example, it has been used to measure transport properties for ionic solutes under steady and nonsteady flow conditions in soils, to monitor water and nitrogen status in the root zone, and to characterize the distribution of water and fertilizers around drippers. Figure 8 illustrates bromide concentrations Copyright  2002 John Wiley & Sons, Ltd. and electrical conductivity measurements in effluent fractions from a soil column compared with ECa measurements using TDR a few centimeters above the column base. The results in Figure 8 are plotted as relative ‘concentrations’ consistent with conventional breakthrough curve format. The excellent correlation of ECa and soil solution ECw shown for this saturated system degenerates under partially saturated conditions, where enhanced tortuosity, variable liquid configuration, and hysteresis complicate the measurement. Calibration models are, therefore, needed to relate measured ECa to the EC or ionic concentration of soil solution because of the dependence of EC on water content as well as on ionic strength and soil geometry. Examples of such models and their application may be found in various papers, including those by Risler et al. (1996), Mallants et al. (1996), Das et al. (1999) and others. Alternative frequency domain analyses Although TDR offers simultaneous and accurate water content and electrical conductivity determination in soils and other porous media, waveform reflections necessary for dielectric constant measurements can be totally attenuated in lossy materials. Factors such as soil texture, salinity, cable length, probe geometry and water content all influence signal attenuation. Attenuated waveforms of varying degree, resulting from a range of solution electrical conductivities, are illustrated in Figure 7. Nadler et al. (1999) found that at field capacity water content in sandy and loamy soils, TDR could be safely used for  measurements up to ECa of approximately 2 dS m1 . TDR applications are, therefore, limited to soils with moderate to low salinity, unless measures are taken to preserve the waveform reflection occurring at the end of the waveguide. Rod coatings have been used successfully to reduce signal attenuation and preserve information needed to evaluate the dielectric constant in highly saline soils. Since these coatings significantly influence the resulting permittivity εb , specific   εb calibration is required for measurements using coated rods, making this a less appealing method (Mojid et al., 1998). Coated rods also make measurement of EC extremely difficult (i.e. requiring extensive calibration over a range of conditions) or ineffective. 149 Hydrol. Process. 16, 141–153 (2002) S. B. JONES, J. M. WRAITH AND D. OR 1.2 Observed Temp-corrected TDR EC 1.0 0.8 0.6 0.4 0.2 Relative concentration 0.0 1.2 EFFLUENT BR 1.0 0.8 0.6 0.4 0.2 0.0 1.2 1.0 EFFLUENT EC 0.8 0.6 0.4 0.2 0.0 0 2 4 6 8 10 Time (d) 12 14 16 Figure 8. Relative EC measured using TDR during steady flow compared with relative bromide concentration and EC in the column effluent fractions. Source: Wraith et al. (1993). Reproduced by permission of Soil Science Society of America The potential benefits of a combination of shorter TDR probes combined with waveform transformations for improving dielectric constant determination at higher levels of electrical conductivity are illustrated in Figure 9. Travel-time analysis provided reliable εb measurement for 10 and 15 cm probes up to ECw ³ 6 dS m1 , whereas, for frequency domain Copyright  2002 John Wiley & Sons, Ltd. analysis, shorter 2 and 3 cm probes extended εb determination by a factor of four to five, up to 24 dS m1 , as shown for a silt loam soil. Conversion of the TDR waveform to the frequency domain provides the frequency-dependent dielectric constant, in addition, other information, such as electrical conductivity, relaxation frequency, and static 150 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING 34 32 30 28 2 cm 3 cm 6 cm 10 cm 15 cm Bulk dielectric constant 26 24 Time domain analysis 22 34 32 Frequency domain analysis 30 28 26 24 22 0 10 20 30 40 50 Soil solution electrical conductivity [dS m−1] Figure 9. Bulk permittivity determined using time domain analysis and frequency domain analysis as a function of probe length and soil solution electrical conductivity. The shaded region indicates the ‘true’ bulk permittivity of the saturated Millville silt loam soil used and high-frequency permittivities, may be extracted using optimization procedures (Heimovaara, 1994; Friel and Or, 1999). Despite the laborious nature of this approach, including fast Fourier transformation of the waveform and fitting of an appropriate model to the transformed scatter function, the procedure has the potential to be automated to make it more amenable to real-time measurements. Outlook In such a short review there are many inevitable omissions of details and complexities of the theory behind TDR. Nevertheless, we have attempted to provide Copyright  2002 John Wiley & Sons, Ltd. potential practitioners with essential elements and an overview of the state-of-practice using TDR. A particularly important advantage of TDR, relative to other methods, is the ability to provide intensive time series measurements, at multiple locations, which are critical to resolution of many hydrologic processes. Concurrent measurement of both  and EC has also provided new research and management opportunities. Since its introduction in the early 1980s, the TDR method has stimulated increased interest in other electromagnetic methods based on different principles ranging from capacitance to frequency-shift sensors. This trend will undoubtedly continue with advances 151 Hydrol. Process. 16, 141–153 (2002) S. B. JONES, J. M. WRAITH AND D. OR in technology and with reduction in costs of electronic components; there are already available several standalone and relatively inexpensive sensors for water content measurement based on dielectric properties. Many of these inexpensive sensors currently have substantial measurement limitations relative to true TDR, however. The TDR method is maturing, as evidenced by the introduction of devices (e.g. Figure 1) specifically designed for hydrological applications. Moreover, the application of alternative methods of analysis, such as frequency domain techniques, provides a means to extend the useful range of utility as well as a potential for extraction of supplementary information concerning water and its interactions with porous media. Some potentially useful applications derived directly from the TDR method include measurement of specific surface area, and in situ determination of water retention properties of field soils. References Baker JM, Allmaras RW. 1990. System for automating and multiplexing soil moisture measurement by time-domain reflectometry. Soil Science Society of American Journal 54: 1–6. Birchak JR, Gardner DG, Hipp JE, Victor JM. 1974. High dielectric constant microwave probes for sensing soil moisture. Proceedings of the IEEE 62: 93–98. Bockris JOM, Devanathan MAV, Muller K. 1963. On the structure of charged interfaces. Proceedings of the Royal Society (London), Series A 274: 55–79. Campbell JE. 1990. Dielectric properties and influence of conductivity in soils at one to fifty megahertz. Soil Science Society of American Journal 54: 332– 341. Chan CY, Knight RJ. 1999. Determining water content and saturation from dielectric measurements in layered materials. Water Resources Research 35(1): 85–93. Dalton FN, Herkelrath WN, Rawlins DS, Rhoades JD. 1984. Timedomain reflectometry: simultaneous measurement of soil water content and electrical conductivity with a single probe. Science 224: 989– 990. Das BS, Wraith JM, Inskeep WP. 1999. Soil solution electrical conductivity and nitrate concentrations in a crop root zone estimated using time-domain reflectometry. Soil Science Society of American Journal 63: 1561– 1570. Dobson MC, Ulaby FT, Hallikainen MT, El-Rayes MA. 1985. Microwave dielectric behavior of wet soil—part II: dielectric mixing models. IEEE Transaction on Geoscience and Remote Sensing GE-23(1): 35–46. Ferre PA, Knight JH, Rudolph DL, Kachanoski RG. 1998. The sample areas of conventional and alternative time domain reflectometry probes. Water Resources Research 34(11): 2971– 2979. Friedman SP. 1998. A saturation degree-dependent composite spheres model for describing the effective dielectric constant of unsaturated porous media. Water Resources Research 34(11): 2949– 2961. Copyright  2002 John Wiley & Sons, Ltd. Friel R, Or D. 1999. Frequency analysis of time-domain reflectometry (TDR) with application to dielectric spectroscopy of soil constituents. Geophysics 64(3): 1–12. Giese K, Tiemann R. 1975. Determination of the complex permittivity from thin-sample time domain reflectometry improved analysis of the step response waveform. Advances Molecular Relaxation Processes 7: 45– 59. Heimovaara TJ. 1994. Frequency domain analysis of time domain reflectometry waveforms: 1. Measurement of the complex dielectric permittivity of soils. Water Resources Research 30(2): 189– 199. Heimovaara TJ, Focke AG, Bouten W, Verstraten JM. 1995. Assessing temporal variations in soil water composition with time domain reflectometry, Soil Science Society of American Journal 59: 689– 698. Jones SB, Or D. 2001. Surface area, geometrical and configurational effects on permittivity of porous media. Journal of Non-Crystalline Solids in press. Mallants D, Vanclooster M, Toride N, Vanderborght J, van Genuchten Mth, Feyen J. 1996. Comparison of three methods to calibrate TDR for monitoring solute movement in undisturbed soil. Soil Science Society of American Journal 60: 747– 754. Mojid MA, Wyseure GCL, Rose DA. 1998. The use of insulated time-domain reflectometry sensors to measure water content in highly saline soils. Irrigation Science 18: 55–61. Nadler A, Gamliel A, Peretz I. 1999. Practical aspects of salinity effect on TDR-measured water content: a field study. Soil Science Society of American Journal 63: 1070– 1076. Or D, Wraith JM. 1999. Temperature effects on soil bulk dielectric permittivity measured by time domain reflectometry: a physical model. Water Resources Research 35(2): 371–383. Reece CF. 1998. Simple method for determining cable length resistance in time domain reflectometry systems. Soil Science Society of American Journals 62: 314– 317. Risler PD, Wraith JM, Gaber HM. 1996. Solute transport under transient flow conditions estimated using time domain reflectometry. Soil Science Society of American Journal 60: 1297– 1305. Robinson DA, Friedman SP. 2000. Parallel plates compared with conventional rods as TDR waveguides for sensing soil moisture. Subsurface Sensing Technologies and Applications 1(4): 137–151. Roth K, Schulin R, Fluhler H, Attinger W. 1990. Calibration of time domain reflectometry for water content measurement using a composite dielectric approach. Water Resources Research 26(10): 2267– 2273. Schaap MG, de Lange L, Heimovaara TJ. 1996. TDR calibration of organic forest floor media. Soil Technology 11: 205– 217. Spaans EJA, Baker JM. 1993. Simple baluns in parallel probes for time domain reflectometry. Soil Science Society of American Journal 57: 668–673. Thorp JM. 1959. The dielectric behaviour of vapours adsorbed on porous solids. Transaction of the Faraday Society 55: 442–454. Topp GC, Davis JL, Annan AP. 1980. Electromagnetic determination of soil water content: measurements in coaxial transmission lines. Water Resources Research 16: 574–582. Weast RC. 1986. In Handbook of Chemistry and Physics, 67th edn. CRC Press: Boca Raton, FL. 152 Hydrol. Process. 16, 141–153 (2002) SCIENTIFIC BRIEFING White I, Zegelin SJ. 1995. Electric and dielectric methods for monitoring soil-water content. In Handbook of Vadose Zone Characterization and Monitoring, Wilson LG, Everett LG, Cullen SJ (eds). Lewis: Ann Arbor; 343– 385. Wraith JM, Or D. 2001. Soil water characteristic determination from concurrent water content measurements in reference porous media. Soil Science Society of American Journal 65(6): in press. Wraith JM, Or D. 1998. Thermo-dielectric estimation of soil specific surface area using TDR. ASA Abstracts. ASA: Madison, WI; 171. Wraith JM, Comfort SD, Woodbury BL, Inskeep WP. 1993. A simplified waveform analysis approach for monitoring solute transport using time-domain reflectometry. Soil Science Society of American Journal 57: 637–642. Wraith JM, Or D. 1999. Temperature effects on time domain reflectometry measurement of soil bulk dielectric constant: experimental evidence and hypothesis development. Water Resources Research 35(2): 361–369. Zegelin SJ, White I, Jenkins DR. 1989. Improved field probes for soil water content and electrical conductivity measurements using time domain reflectometry. Water Resources Research 25(11): 2367– 2376. Copyright  2002 John Wiley & Sons, Ltd. 153 Hydrol. Process. 16, 141–153 (2002)