|
|
||||||||
a Department of Agronomy, Iowa State University, Ames, IA 50011-1010 USA
b Greenhouse and Processing Crops Research Centre, Agriculture and Agri-Food Canada, Harrow, ON Canada N0R 1G0
rhorton{at}iastate.edu
| ABSTRACT |
|---|
|
|
|---|
, and water content,
, is demanded in many disciplines. Time domain reflectometry (TDR) has been routinely used to measure water content and electrical conductivity of soil. Previous efforts to combine TDR probes with porous materials functioning as tensiometers were successful, but these probes were still constrained by characteristics of tensiometers, such as the need to supply water and measuring ranges
> -85kPa. A new
TDR probe was developed to overcome shortcomings of the previous work. A portion of the TDR rod was embedded in a dental plaster (gypsum), whose matric potential equilibrated with surrounding soil. The rest of the TDR rod was inserted into the soil. The TDR technique was used to determine dielectric constants,
, of both the gypsum and the soil. The new
TDR probes were tested in clay loam soil using a pressure-plate apparatus to produce
relationships of the gypsum and the soil for -1000
-10 kPa. Changes in
of the gypsum corresponded well to applied pressures for -1000
-30 kPa, but
values did not noticeably change for
> -30 kPa. Values of
of the soil corresponded well to the whole
range tested. The new probes accurately measure
and
of soil when soil water content gradually decreases or increases. The newly developed
TDR probe requires no more maintenance than ordinary TDR probes and requires no additional instrumentation.
Abbreviations: TDR, time domain reflectometry
| INTRODUCTION |
|---|
|
|
|---|
(m3 m-3), and matric potential,
(kPa)is one of the most important parameters to study water flow in soil. A pressure-plate apparatus, combined with the oven-drying soil method, is popularly used to establish water characteristic curves in the laboratory. However, in situ determination of water characteristic curves is difficult. Measuring soil water content may be relatively easy in both destructive (e.g., sampling soils) and nondestructive (e.g., TDR, neutron scattering method) manners. Matric potential of soil is often measured with a tensiometer in situ; however, its measuring range is limited to
> -85 kPa (Cassel and Klute, 1986). Thermocouple psychrometry can be used for a wide range of
but is very sensitive to temperature changes (Rawlins and Campbell, 1986); thus, it may not be suitable for routine field measurements. Measuring water, thermal, or electrical properties of a constructed porous medium equilibrated with surrounding soil is another attempt to indirectly measure the matric potential of soil. For specific ranges of water potential, the heat-dissipation method, the filter-paper method, or the gypsum block electrical-resistance method may be used in the laboratory and the field (Campbell and Gee, 1986).
Time domain reflectometry is routinely used to measure water content and electrical conductivity both in the laboratory and the field (Dalton et al., 1984; Nadler et al., 1991; Noborio et al., 1994; Heimovaara et al., 1995). For simultaneous measurement of
and
of soil with a TDR probe, Baumgartner et al. (1994) and Whalley et al. (1994) attached porous materials functioning as tensiometers to the end of hollow electrodes of the TDR probe. However, their probe configuration has the same constraints as tensiometers, that is, the need to supply water to tensiometers and the limited measuring range for
> -85 kPa.
To estimate
values by measuring the dielectric constant,
, or water content of an equilibrated porous material, TDR techniques have been applied to a commercial product (e.g., Equitensiometer, Delta-T Devices, Cambridge, England) and to a TDR probe in ceramic discs described by Or and Wraith (1999a). When both
and
need to be measured, an additional probe is required to measure
. There is a need to develop a probe that simultaneously measures wide ranges of
and
in situ.
In this paper, we present a design of a new TDR probe that simultaneously measures
and
. A portion of the TDR probe is embedded in a porous material, and the rest is inserted into a surrounding soil. This unique configuration of the probe enables the simultaneous measures of
for the porous material and
for the soil, side by side, in almost the same sampling volume. The new
TDR probes are tested in clay loam soil.
| Materials and methods |
|---|
|
|
|---|
), for a two-rod type transmission line can be approximated (Kraus, 1984) using Eq. [1]:
![]() | (1) |
is the dielectric constant of a material surrounding the transmission line, s is the spacing of the rods, and d is the radius of the rods. To distinguish reflections from the interface between different diameters and spacings of rods, the impedance differences should be large enough in a full range of soil water contents, that is, dielectric constants. The two-rod transmission line or probe consisted of two stainless-steel rods (1.6-mm diam. and 25 mm apart) in which a portion was covered with copper tubes with larger outer diameters (3.2-mm diam., 50 mm long, and 5 mm apart) (Fig. 1) . The copper tubes were soldered to the stainless-steel rods at the both ends of the copper tubes. The copper tubes were then embedded in dental plaster (Lab Plaster, Bayer Corp., South Bend, IN) following Phene et al. (1971). Lab Plaster's main constituent is gypsum. The powder plaster was mixed with deionized water in a ratio of 47 mL of water to 100 g of the Lab Plaster. The slurry of the plaster was poured into a plastic cast in which the copper portion of the two rods was placed at the center of the cast.
|
10 mm. Thus,
measurement of the gypsum with the configuration shown in Fig. 1 should be accurate, and the probe should not require a wire screen as did Wraith and Baker (1991) to confine the electromagnetic energy.
According to Eq. [1], the stainless-steel portion of the transmission line in dry and wet soil yield Z = 238
and 75
, respectively, whereas the copper portion in dry and wet gypsum results in Z = 58
and 25
, respectively. The dielectric constant of dry and wet soil was assumed to be 3 and 30 (Topp et al., 1980), respectively, and that of dry gypsum was assumed to be 5.6 (Curtis and Defandorf, 1929). The dielectric constant of wet gypsum was assumed to be the same as that of wet soil. The diameters and spacings of rods were selected to provide distinguishable diffraction of TDR waves at the interface between gypsum and soil for all combinations of water contents of gypsum and soil. A 75-
coaxial cable (RG-187/A, Alpha Wire, Elizabeth, NJ) was connected to the end of the copper tubes. The 75-
coaxial cable provided a distinguishable reflection from the beginning of the probe in the full range of water contents of gypsum, especially when the gypsum was dry. Two
TDR probes were constructed and tested.
Calibration
The
TDR probes were embedded in Harps clay loam (fine-loamy, mixed, superactive, mesic Typic Caliaquoll; 37% sand, 35% silt, and 28% clay) in a pressure-plate apparatus. A layer
1.3 cm thick of air-dried Harps clay loam was placed in a PVC cylinder (15.2-cm diam. and 5.0 cm high) on a pressure plate, then two probes were horizontally placed on the soil. Additional air-dried soil filled gaps between the probes and the cylinder. The extra 1.3-cm thickness of soil covered the probes on top. Using Eq. [4] and [5] of Petersen et al. (1995), who modified Knight's work (1992), the 1.3-cm-thick soil was sufficient to involve 95% of the total electromagnetic energy provided by a TDR cable tester. The soil and the probes were saturated with deionized water for 1 d and placed in a pressure chamber. Nine different pressures were applied to the system in a constant temperature room (20°C). After the soil and gypsum were assumed to be equilibrated with each applied pressure (equilibration times ranged from 2 to 16 d), the pressure chamber was opened, and TDR waveforms were collected using a 1502 TDR cable tester (Tektronix Inc., Beaverton, OR) connected to a 21X datalogger (Campbell Scientific, Logan, UT) for control and data storage. Acquired TDR waveforms were transferred to a computer and analyzed using a procedure similar to that of Baker and Allmaras (1990). Waveform acquisition from the cable tester was duplicated for each pressure applied. Additional soil columns (4.75-cm i.d. and 5 cm high) were placed on the same pressure plate for determining equilibrated soil water content for each applied pressure.
Dielectric constants of gypsum and soil were determined (Baker and Allmaras, 1990) in Eq. [2] as
![]() | (2) |
is the dielectric constant of gypsum or soil, La is an apparent probe length on a cable tester (m), and L is a probe length (m). Apparent probe lengths for gypsum and soil were represented by La,g and La,s, respectively, as in Fig. 2
. The TDR-measured dielectric constant,
, of the gypsum was related to matric potential,
(kPa), with an equation (Eq. [3]) similar to that of van Genuchten (1980):
![]() | (3) |
s and
r are dielectric constants at saturation and residual water contents, respectively, and
, n, and m are calibration constants. The relationship between the TDR-measured dielectric constant,
, and the volumetric water content of soil,
(m3 m-3), was established (Eq. [4]) (Yu et al., 1997) with:
![]() | (4) |
|
TDR probe during desorption was evaluated in a pressure chamber. The soil and the probe were saturated with deionized water for 1 d on a pressure plate, then placed in a pressure chamber. A 100-kPa external pressure was applied, and temporal changes in TDR waveforms were measured at 1-h to 1-wk intervals. Following desorption, deionized water was sprinkled to near saturation on the surface of the same soil and probe. Temporal changes in TDR waveforms were again collected to evaluate response time during sorption without applying external pressures. | Results and discussion |
|---|
|
|
|---|
= -10, -100, and -1000 kPa in Fig. 2. Reflections from the beginning and end of the probe, in addition to the reflection from the interface between gypsum and soil, were consistently distinguishable in the whole pressure range we tested. Waveforms from Probe 2 were very similar to Probe 1 (data not shown).
Changes in the dielectric constant of the gypsum and the soil corresponding to external pressures applied are shown in Fig. 3 . Dielectric constants of the gypsum and the soil were determined using Eq. [2] with apparent probe lengths La,g for the gypsum and La,s for the soil as shown in Fig. 2. The observed
relationship for the gypsum was almost identical for the two
TDR probes. The
relationship for the gypsum was established using Eq. [3] with data from these two probes. Fitted parameters were
r = 3.158,
s = 21.452,
= 0.0207, n = 3.150, and m = 0.228. The correlation coefficient was r2 = 0.998. If the
relationship for the gypsum differed from soil to soil, it would be necessary to confine the electromagnetic energy by enclosing the gypsum with a wire screen (e.g., Wraith and Baker, 1991). The observed
for the soil was similar for the two probes, but differed by as much as
= 0.85, which is equivalent to
= 0.016 m3 m-3. This difference may be within the acceptable error range. Changes in
for the gypsum were sensitive enough to detect a change for -1000
-30 kPa, but almost insensitive for
> -30 kPa. These characteristics of the gypsum used were similar to those of the gypsum reported by Bourget et al. (1958) and those of the ceramic discs used by Or and Wraith (1999a). In contrast, a nylon block (Bourget et al., 1958) and a porous material used for a heat-dissipation sensor (Reece, 1996) are sensitive at -1
-100 kPa, but lose sensitivity at
> -100 kPa. A porous material having a semi-log linear relationship between
and
in a wide range of matric potential, perhaps 0 to -1500 kPa, may be ideal.
|
TDR probes were embedded in quartz sand for calibration, the gypsum did not respond well to desorption of the sand (data not shown). Or and Wraith (1999a) reported a similar limitation of their TDR-based matric potential sensor in coarse-textured soils. This phenomenon might result from restricted water flow at the interface between the smooth surface of a porous sensor material and coarse soils (Gardner, 1986). Gardner (1986) suggested coating the porous sensor material with diatomaceous earth to reduce this phenomenon.
Advantages of
measurement using TDR may be extended to the
calibration of the gypsum because there is little effect of temperature and salinity on
or
determination (Topp et al., 1980). Unlike other methods for measuring
, such as thermocouple psychrometry and the gypsum block-resistance method (Campbell and Gee, 1986), we can expect temperature independence in the
relationship because Halbertsma et al. (1995) reported that a fine-textured soil showed little effect of temperature on TDR-measured
. Others, however, reported temperature dependency of
in fine-textured soils (Pepin et al., 1995; Wraith and Or, 1999b). Or and Wraith (1999b) proposed a mechanistic model for the temperature dependency of soil
. Thus, we may need temperature compensation for the gypsum, which has tight pore-size distribution, when the probe is exposed to large temperature differences such as occur in surface soil. Although salinity effects on TDR-measured
are not completely understood (Ren et al., 1999), the
relationship should be consistent when the calibration is made under fixed conditions, that is, the gypsum block is always moistened with a saturated solution of dissolved gypsum (Gardner, 1986).
Hysteresis in the gypsum, however, may influence the calibration. Tanner and Hanks (1952) and Bourget et al. (1958) found hysteresis when they measured electrical resistance between electrodes in a gypsum block as a function of matric potential. Although hysteresis between TDR-measured
and
was negligible (Topp et al., 1980; Horino and Maruyama, 1993), there may be hysteresis between
and
in the gypsum because the hysteresis between
and
in other porous media (e.g., soil) has been evident (Royer and Vachaud, 1975; Watson et al., 1975).
A calibration curve for the clay loam was made by fitting the volumetric water content and the TDR-measured dielectric constant to Eq. [4]. Calibration constants were then determined as a = -0.0340, b = 0.0458, and c = 0.718, with r2 = 0.987 (Fig. 4A) . The calibration curve was determined using data collected by Probes 1 and 2. Water content measured using TDR with Eq. [2] and [4] agreed well with gravimetrically determined
, for
> 0.12 m3 m-3. For 0.05 <
0.12 m3 m-3, however, TDR overestimated
. This overestimation may not be attributed to the special configuration of our probe design because Ren et al. (1999) reported the same trend using a 4-cm long three-rod probe in the same soil. When the general calibration curve (Topp et al., 1980) was used instead of Eq. [4], the middle range of water contents was well estimated, but performance was poor near saturation and in dryer regions (Fig. 4A). Using calibration curves for matric potential (Eq. [3], Fig. 3) and for water content (Eq. [4], Fig. 4A) estimated using the
TDR probes, water characteristic curves for the clay loam were comparable to those determined by the conventional pressure-plate apparatus (Fig. 4B). Deviation from the data collected by the conventional method was found at
> -20 kPa and 500 <
< -200 kPa, due to calibration errors for the
relationship of the gypsum and the
-
relationship of the clay loam, respectively.
|
for the gypsum and the soil during desorption and absorption are shown in Fig. 5A and B
, respectively. During desorption at
= -100 kPa (Fig. 5A), the soil
started decreasing just after pressure was applied and reached the first plateau in 40 h, then decreased again to the equilibrium value. The gypsum
increased slightly in the first few hours, then started decreasing as with the soil (similar to Or and Wraith, 1999a). The response time of the gypsum during desorption corresponded well to that of the soil.
|
gradually decreases by evapotranspiration and increases by slow water infiltration. However,
measurement under conditions of quick changes in water status near the soil surface due to rain or irrigation may be less reliable. | Conclusions |
|---|
|
|
|---|
TDR probe that simultaneously measures matric potential and water content of soil was developed and tested in the laboratory. Like previously reported devices using gypsum as a porous medium, our new probes showed the similar response of
changes with an effective range of
-30 kPa. Water content measured by these probes agreed with gravimetrically determined
using a soil-specific calibration. The response time was similar to other devices using gypsum as a porous medium. This new TDR probe can be used to simultaneously monitor
and
in the field. More research is needed to find (i) ideal porous materials with a wide range of water retention and a fast response to the water status of the surrounding soil, and (ii) temperature and hysteresis effects on the
relationship of the porous materials.Wraith Or 1999 | ACKNOWLEDGMENTS |
|---|
| NOTES |
|---|
|
|
|---|
Received for publication September 21, 1998.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
X. L. Xin, F. A. Xu, J. B. Zhang, and M. X. Xu A New Resistance Sensor for Monitoring Soil Matric Potential Soil Sci. Soc. Am. J., May 16, 2007; 71(3): 866 - 871. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Persson, J. M. Wraith, and T. Dahlin A Small-Scale Matric Potential Sensor Based on Time Domain Reflectometry Soil Sci. Soc. Am. J., February 27, 2006; 70(2): 533 - 536. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. A. Robinson, D. A. Robinson, S. B. Jones, J. M. Wraith, D. Or, and S. P. Friedman A Review of Advances in Dielectric and Electrical Conductivity Measurement in Soils Using Time Domain Reflectometry Vadose Zone J., November 1, 2003; 2(4): 444 - 475. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. M. P. Vaz, J. W. Hopmans, A. Macedo, L. H. Bassoi, and D. Wildenschild Soil Water Retention Measurements Using a Combined Tensiometer-Coiled Time Domain Reflectometry Probe Soil Sci. Soc. Am. J., November 1, 2002; 66(6): 1752 - 1759. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Wraith and D. Or Soil Water Characteristic Determination from Concurrent Water Content Measurements in Reference Porous Media Soil Sci. Soc. Am. J., November 1, 2001; 65(6): 1659 - 1666. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Journal of Environmental Quality |
||||