Published in Vadose Zone Journal 2:650-654 (2003)
© 2003 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SPECIAL SECTION - ADVANCES IN MEASUREMENT AND MONITORING METHODS
Laboratory Evaluation of a Commercial Dielectric Soil Water Sensor
Bobbie McMichaela and
Robert J. Lascano*,b
a USDA-ARS, 3810 4th Street, Lubbock, TX 79415
b Texas A&M University, 3810 4th Street, Lubbock, TX 79415
* Corresponding author (r-lascano{at}tamu.edu)
1 Mention of a trade name does not constitute a guarantee or warranty of the product by the United States Department of Agriculture and does not imply its approval to the exclusion of other products that may also be available. 
Received 12 March 2003.
 |
ABSTRACT
|
|---|
Development of management strategies for efficient water utilization of crop production requires measurements of changes in soil water content on a dynamic basis. Many of the methods currently used for measuring these changes are destructive, slow, or relatively expensive for large-scale investigations. A commercially available, low-cost, nondestructive soil moisture sensor for measuring changes in soil volumetric water content (VWC) on the basis of changes in the dielectric constant of the soil water was evaluated under laboratory conditions for two soil series (Amarillo fine sandy loam [fine-loamy, mixed, superactive, thermic Aridic Paleustalfs] and Pullman clay loam [fine, mixed, thermic Torretic Paleustolls]) and a potting material across a wide range of water contents. Probes were placed in containers filled with deionized water and soil. Containers with Amarillo fine sandy loam were placed in a programmable temperature chamber and subjected to a series of changes in both temperature and VWC. Containers with Pullman soil and potting material were only subjected to changes in VWC at a constant temperature. Probe output at a constant temperature between air dry and a VWC of 0.25 m3 m-3 was linear for the Pullman soil and potting material and nonlinear for the Amarillo soil. When the Amarillo soil temperature varied between 15.9 and 39.1°C-1 at a constant VWC, probe output changed the equivalent of 0.10 m3 m-3. The temperature sensitivity was 0.5 mV °C-1 for air-dry and about 5 mV °C-1 for wet Amarillo soil. We conclude that probe output is soil specific and, given the nonlinear response to increasing water content on some soils and sensitivity to temperature, will require soil-specific calibration equations.
Abbreviations: TDR, time domain reflectometry • VWC, volumetric water content
|
INTRODUCTION
|
|---|
THE MEASUREMENT OF changes in soil VWC is needed to describe plant responses to soil drying, characterizing soil water availability and soil water evaporation. For the most part, these measurements are conducted under field conditions, and methods used range from gravimetric methods to techniques such as time domain reflectometry (TDR) (Topp et al., 1980). In addition, tensiometers are widely used to measure the soil matric potential of relatively wet soils and are limited to potentials greater than -800 kPa because of cavitation of the water column inside the tensiometers. Some methods such as neutron attenuation, for example, allow for measurements to be made at various depth increments in the soil (Hignet and Evett, 2002), while other techniques such as TDR provide a single measurement integrated across a range of depths or a single depth if probes are oriented horizontally in the soil.
The concept behind the more recent methods deals with the relationships between changes in the soil dielectric constant as a function of soil VWC and the ability to measure these changes. The TDR method, for example, is based on the determination of the changes in the velocity of an electromagnetic pulse sent through a probe (wave-guide) inserted into the soil. Differences in time required for the pulse to traverse the length of the wave-guide and return depend on the soil dielectric constant and consequently on the soil VWC.
Another technique that depends on the changes in the soil dielectric constant to measure water content is the capacitance method. Here a capacitor (probe) is subjected to a specific voltage and the charge time is measured. The charge time is a function of the capacitance of the probe, which is directly related to the dielectric constant of the medium (soil).
Both TDR and capacitance methods depend on changes in the soil dielectric constant to measure soil VWC. Therefore, ease of use, and other factors affecting the output of the instruments (e.g., temperature, salinity) become considerations in the choice of methods (Wraith and Or, 1999; Or and Wraith, 1999; Baumhardt et al., 2000).
A low-cost capacitance probe is currently being manufactured and is commercially available for use in a wide range of soil types. The objective of this study was to evaluate the effect of changes in water contents for two soil series and a potting material on the output and sensitivity of these probes under controlled laboratory conditions. In our study we did not compare the capacitance probe with TDR. In addition, the combined effect of changes in temperature and water content was evaluated on a single soil.
|
MATERIALS AND METHODS
|
|---|
Description of Soil Moisture Probe
The ECH2O dielectric Aquameter Model EC20 (0.2-m length) probe (Decagon Devices, Pullman, WA) was used as the soil moisture sensor.1 Each probe was equipped with a standard 3-m (10-ft) cable and an adaptor cable for connection to a datalogger (Model CR-21X, Campbell Scientific, Logan, UT).
To evaluate the use of ECH2O probes to measure soil VWC and their sensitivity to temperature fluctuations, two types of studies were performed, all under laboratory conditions. The first study evaluated the output sensitivity of ECH2O probes to changes in temperature from about 16 to 39°C, with the probes fully immersed in water. The second study was similar to the first one, except that probes were buried in different soils and a potting material packed in containers and subjected to changes in VWC at a constant temperature. In addition, ECH2O probe output subjected to changes in both VWC and temperature was evaluated on one soil type.
In preliminary tests with the ECH2O probes, we determined that the sensor sensitivity along its length was independent of their vertical orientation in containers filled with soil. To test this we placed two ECH2O probes with one probe inserted in such a way that its cable connector was flush with the soil surface and the other probe was inverted with its cable connector flush with the bottom of the container. We compared the output from both probes for various soil VWC values, ranging from air dry to 0.25 m3 m-3, and found no significant statistical differences (i.e., analysis of variance and mean differences at a p = 0.05) among probe readings. Therefore, the orientation of the probes for the remainder of our studies was as shown in Fig. 1, with the cable connector flush with the soil surface for both ECH2O probes.
View larger version (58K):
[in this window]
[in a new window]
|
Fig. 1. Experimental setup of ECH2O probes and thermocouples inside a PVC cylinder filled with soil (not to scale).
|
|
Study 1: Probes in Water
Four ECH2O probes were suspended in a 5.0-L beaker. A styrofoam cover was fashioned to fit over the top of the beaker to minimize water evaporation with four slits cut into the cover allowing the probes to be suspended inside the beaker. The beaker was filled with deionized water, ensuring that the sensor portion of each probe was fully covered. To measure water temperature, three thermocouples (type-T) were suspended near the surface, middle, and bottom of the beaker. A datalogger (Model CR-21X, Campbell Scientific) was used to record temperature and ECH2O output every 30 s with an average output every 5 min. We used the manufacturer's recommended excitation voltage of 2.5 V when reading the ECH2O probes.
Both beaker and datalogger were placed inside a temperature programmable chamber. The initial chamber temperature was set to approximately 15°C and allowed to equilibrate for at least 2 h. Thereafter, the initial temperature was maintained for 4 h and increased by 5°C every 2 h to approximately 40°C. Once the final temperature was reached, the cycle was reversed by decreasing the temperature to approximately 15°C in 5°C increments every 2 h. To assure repeatability of measurements these experiments were repeated three times.
Study 2: Probes in Soil and Potting Material
The soils used in this study were two Amarillo fine sandy loams sampled at Lubbock and Lamesa, TX and a Pullman clay sampled at Halfway, TX. The Amarillo soil sampled in Lubbock, hereafter referred to as Amarillo-1, was obtained at the USDA-ARS Plant Water Stress Laboratory in Lubbock and was 19% clay and 68% sand. The second Amarillo soil, hereafter referred to as Amarillo-2, was obtained at the Lamesa Agricultural Research Farm of Texas A&M University and was 16% clay and 78% sand. The Pullman soil was obtained at the Halfway Texas A&M University Agricultural Research Farm and was 39% clay and 37% sand. The particle-size distribution was measured using the hydrometer method on triplicate samples (Gee and Bauder, 1986). Additional information on the physical and hydraulic properties of these soils is given by Baumhardt et al. (1995). These soils represent two of the major agricultural series in the Texas High Plains and are soils with differing particle-size distribution. In addition, we also used and evaluated a commercially available potting material known as fritted clay, a granular material made by firing coarsely-milled, dry clay in a rotary kiln mostly in the form of granules 1 to 2 mm in size (Van Bavel et al., 1978).
Before filling the containers, soils were air dried and passed through a 2-mm sieve. Polyvinyl chloride cylinders (0.45 m deep by 0.15 m in diameter) were used as containers. Approximately 70 mm of pea gravel was placed in the bottom of each cylinder before inserting the ECH2O probes. Seven centimeters of soil was placed over the gravel, and the ECH2O probes (two for each of two containers) were inserted (Fig. 1). Two ECH2O probes were centered in each of the two containers separated by approximately 50 mm. Thermocouples (type-T) were placed between the two probes at three locations (top, middle, and bottom; see Fig. 1) in the containers. As soil was added, containers were lightly tapped to consolidate soil particles trying to achieve a relatively uniform bulk density. The probes were covered with soil until only the cable extended above the soil surface. The tops of the containers were covered with a plastic wrap and secured with rubber bands to minimize water evaporation.
Mass of the PVC containers, with ECH2O probes, pea gravel, and thermocouples with and without soil and potting mixture were recorded (±0.1 g), and corresponding dry soil bulk density calculated. For each container, the amount of water needed to increase the soil VWC by approximately 0.05 m3 m-3 was gravimetrically calculated and typically was around 0.250 kg for the three soils used. Water was added to the top of the PVC containers. Containers with soil were placed in a growth chamber and the temperature set to 25°C. A datalogger (Model CR-21X, Campbell Scientific) was used to record data every 30 s with an average output every 5 min at an excitation voltage of 2.5 V. For the Amarillo-1 soil at each VWC (air dry to 0.25 m3 m-3) the chamber temperature was cycled from approximately 15 to 40°C, as described in Study 1. These measurements were repeated four to six times. To measure the distribution of water in several containers the soil VWC was measured gravimetrically from soil samples collected in 0.07-m increments.
|
RESULTS AND DISCUSSION
|
|---|
Probes in Water
As an example, output from an ECH2O probe as a function of temperature when immersed in deionized water is shown in Fig. 2, and mean output for two probes is given in Table 1. In our experiments, using four to six ECH2O probes, statistical analysis indicated no significant differences among probes. Therefore, only results from one of several probes used are shown as examples for the soil and potting material studies. Measurements indicated that as the water temperature increased from 13.3 to 39.2°C, output from the ECH2O probes increased from a minimum of 1029 and 1038 mV to 1057 and 1066 mV, respectively (a 28-mV difference shown in Table 1). The output of the probes subsequently decreased as the temperature was lowered to near starting levels (Fig. 2). In water, the effect of changing the temperature was 1.1 mV °C-1 and was uniform across all ECH2O probes tested. There was also a hysteretic effect in the probe output as the water temperature was cycled from low to high and then to low temperature as shown in Fig. 2. We have no explanation for this temperature dependence.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2. Output of an ECH2O probe in deionized water as a function of increasing and decreasing temperature. The arrows indicate the order of ascending or descending temperature.
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Mean, standard deviation, maximum, and minimum of two ECH2O probes (mV) and soil temperature measured when the temperature was varied at a soil VWC of 0.15 and 0.19 m3 m-3 in an Amarillo-1 soil and in water.
|
|
Probes in Soil and Potting Material
As an example, the gravimetrically measured distribution of soil water in three containers with Amarillo-1 soil as a function of depth is shown in Fig. 3. The mean soil VWC values for the containers were 0.038, 0.198, and 0.247 m3 m-3. The VWC was calculated using the measured values of gravimetric water content and bulk density. As expected, the largest discrepancy was near the surface; however, the deviations from the mean were <0.02 m3 m-3. Since the distribution of the VWC in the containers was relatively uniform, our results suggest that in our tests, and for a wide range of soil water contents, output from the ECH2O is proportional to the sensor portion along the overall length of the probes tested. For example, in the drier container (Fig. 3) the length of the probe is exposed to soil VWC that range from 0.02 to 0.06 m3 m-3, yet the probe output is proportional to the mean soil VWC in the container. However, our tests are for uniform soil in a container, and the effect of different soil horizons on ECH2O output remains to be tested.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 3. Measured soil volumetric water content (VWC) as a function of depth at an average soil VWC of 0.038, 0.198, and 0.247 m3 m-3, respectively, on three separate cylinders.
|
|
Figure 4 illustrates output from an ECH2O probe as a function of time at different soil VWC in an Amarillo-1 soil at a constant temperature of 26.4°C and a temperature cycling between 15.5 and 39.1°C at 0.15 m3 m-3 soil VWC. Output for the probe was 394.1 ± 4.0 mV in air-dry soil, which corresponds to a measured soil VWC of 0.016 ± 0.002 m3 m-3. Probe output increased to about 825 mV at a soil VWC of 0.25 m3 m-3. When the temperature varied between 15.5 and 39.1°C at a constant soil VWC of 0.15 m3 m-3, probe output varied by about 120 mV (Table 1, Fig. 4). This variation represents a change in soil VWC of approximately 0.10 m3 m-3. The results shown in Fig. 4 clearly show that for soil water contents <0.15 m3 m-3 the ECH2O probes respond in a step-function; that is, as soil VWC increases, the probe output increases linearly, and the response is instantaneous. However, for soil VWC >0.15 m3 m-3, probe output is nonlinear and the effect of changing temperature at a constant soil VWC makes the determination of soil water content difficult. For example, when soil VWC was 0.19 m3 m-3, the variation in the output of the probes was approximately 132 mV across a temperature range from 14.7 to 39.5°C, which also represented a change of approximately 0.10 m3 m-3 (Table 1).
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4. Output of an ECH2O probe as a function of day of year (DOY, 2002) and different soil volumetric water content (VWC) at a constant soil temperature of 26.4°C. At a soil VWC of 0.15 m3 m-3 the soil temperature varied between 15.5 and 39.1°C. The vertical dashed lines represent the time when water was added to the PVC cylinder.
|
|
ECH2O output as a function of soil VWC for the Amarillo-1 soil is shown in Fig. 5. At each level of soil VWC the vertical bars represent the range of individual data points of ECH2O output when the temperature varied between 14.5 and 39.0°C, with a mean of 26.4°C. Again, this result shows the linear response for soil water contents <0.15 m3m-3 and its nonlinearity for larger values, indicating that temperature affects the probe output and cannot be ignored. Echo output for Amarillo-2, Pullman soils, and fritted clay at a constant temperature of 24.5°C is shown in Fig. 6a through 6c. In these figures the vertical bars are individual data points that represent the range of ECH2O output at a constant temperature. The response of the Amarillo-2 soil (Fig. 6a) is similar to that of the Amarillo-1 (i.e., linear response for soil VWC of <0.15 m3 m-3 and nonlinear for larger values of soil VWC; Fig. 6a). However, the response of the ECH2O probes in a clay soil (Fig. 6b) and a potting material (Fig. 6c) was linear as soil VWC increased between air dry and 0.25 m3 m-3. These results clearly show that the response of the ECH2O probes is soil specific and thus requires individual calibration between ECH2O output and soil VWC. The equations shown in Fig. 5 and Fig. 6a through 6c were obtained using step-wise multiple regression, and the equation with the highest coefficient of multiple determination (R2) was selected (SAS Institute, 1998). These equations have no physical meaning other than to show that the trend between ECH2O output and VWC at different temperatures was not always linear.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5. ECH2O output as a function of soil volumetric water content (VWC), where temperature varied between 14.5 and 39.0°C at each water level. The vertical bars at each VWC represent the distribution of individual data points for the temperature variation.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 6. ECH2O output as a function of soil volumetric water content (VWC) for two soils and a potting material all at a constant temperature of 24.5°C: (a) Amarillo-2, (b) Pullman, and (c) fritted clay. The vertical bars represent the distribution of individual data points at each soil VWC.
|
|
The sensitivity of the ECH2O probes to temperature as a function of soil VWC measured in an Amarillo-1 soil is shown in Fig. 7. These results summarize all the combinations of varying temperature and soil VWC in all the ECH2O probes tested in this soil. These results show that the sensitivity to temperature increases linearly from 0.5 mV °C-1 for air-dry soil to about 4.5 mV °C-1 for a VWC of 0.10 m3 m-3. Further, as the VWC increases from 0.10 to 0.20 m3 m-3, the sensitivity increases slightly to 5.5 mV °C-1 and then decreases to 4.5 mV °C-1 at a VWC of 0.25 m3 m-3. The effect of temperature on ECH2O output cannot be ignored and needs to be considered; otherwise, the determination of soil VWC with these probes will be difficult, especially in the wetter range. We recommend that if instantaneous values of soil VWC are needed, it will be necessary to also measure the soil temperature.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 7. Temperature sensitivity of ECH2O probes output as a function of changes in soil volumetric water content (VWC) in a fine sandy loam soil.
|
|
|
CONCLUSIONS
|
|---|
These results illustrate that these sensors require soil-specific calibration equations and that in sandier soils the probes have limited application to measure soil VWC above 0.15 m3 m-3 even if temperature data were available. However, their response on a clay loam soil and potting material was linear with increasing soil VWC. In sandy soils, the use of these sensors is limited to drier soil conditions and cannot be used near the soil surface where both temperature and water content would have the largest fluctuations. It is also suggested that users of the probes perform their own calibration for their particular application. The use of the probes may be adequate as a low-cost alternative to measure soil water content for many agricultural cropping system applications, even with the observed limitations and restrictions.
|
REFERENCES
|
|---|
- Baumhardt, R.L., R.J. Lascano, and S.R. Evett. 2000. Soil material, temperature, and salinity effects on calibration of multisensor capacitance probes. Soil Sci. Soc. Am. J. 64:1940–1946.[Abstract/Free Full Text]
- Baumhardt, R.L., R.J. Lascano, and D.R. Krieg. 1995. Physical and hydraulic properties of Pullman and Amarillo soil on the Texas South Plains. Tech. Rep. 95-1.Texas A&M University Res. and Ext. Center, Lubbock/Halfway.
- Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 383–423. In A. Klute (ed.) Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- Hignett, C., and S.R. Evett. 2002. Neutron thermalization. p. 501–521. In J.H. Dane and G.C. Topp (ed.) Methods of soil analysis. Part 4. SSSA, Madison, WI.
- Or, D., and J.M. Wraith. 1999. Temperature effects on soil bulk dielectric permittivity measured by time domain reflectometry: A physical model. Water Resour. Res. 35:371–383.
- SAS Institute Inc. 1998. StatView reference. SAS Inst., Cary, NC.
- Topp, G.L., J.L. Davis, and A.P. Annan. 1980. Electromagnetic determination of soil water content: Measurement in coaxial transmission lines. Water Resour. Res. 16:574–582.
- Van Bavel, C.H.M., R.J. Lascano, and D.R. Wilson. 1978. Water relations of fritted clay. Soil Sci. Soc. Am. J. 42:657–659.[Abstract/Free Full Text]
- Wraith, J.M., and D. Or. 1999. Temperature effects on soil bulk dielectric permittivity measured by time domain reflectometry: Experimental evidence and hypothesis development. Water Resour. Res. 35:361–369.
This article has been cited by other articles:
|
|
|
|
 
D. A. Robinson, C. S. Campbell, J. W. Hopmans, B. K. Hornbuckle, S. B. Jones, R. Knight, F. Ogden, J. Selker, and O. Wendroth
Soil Moisture Measurement for Ecological and Hydrological Watershed-Scale Observatories: A Review
Vadose Zone J.,
February 25, 2008;
7(1):
358 - 389.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Fares, H. Hamdhani, and D. M. Jenkins
Temperature-Dependent Scaled Frequency: Improved Accuracy of Multisensor Capacitance Probes
Soil Sci. Soc. Am. J.,
May 16, 2007;
71(3):
894 - 900.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. R. Evett and G. W. Parkin
Advances in Soil Water Content Sensing: The Continuing Maturation of Technology and Theory
Vadose Zone J.,
November 11, 2005;
4(4):
986 - 991.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. M. Blonquist Jr., S. B. Jones, and D. A. Robinson
Standardizing Characterization of Electromagnetic Water Content Sensors: Part 2. Evaluation of Seven Sensing Systems
Vadose Zone J.,
November 11, 2005;
4(4):
1059 - 1069.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
