Published online 16 May 2007
Published in Soil Sci Soc Am J 71:866-871 (2007)
DOI: 10.2136/sssaj2006.0195
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SOIL PHYSICS
A New Resistance Sensor for Monitoring Soil Matric Potential
X. L. Xin,
F. A. Xu,
J. B. Zhang* and
M. X. Xu
State Experimental Station of Agro-Ecosystem in Fengqiu; State Key Lab. of Soil and Sustainable Agriculture Inst. of Soil Science, Chinese Academy of Sciences, Nanjing 210008, People's Republic of China
* Corresponding author (jbzhang{at}issas.ac.cn).
 |
ABSTRACT
|
|---|
The electrical resistance sensor is an appropriate and inexpensive tool for measuring soil matric potential (
). These sensors are widely used in irrigation agriculture. One drawback of the resistance sensors, however, is that the measurements often are not accurate across a wide range of moisture conditions. The objective of this study was to design a resistance sensor that can measure accurately and precisely across a wide range of moisture conditions, thereby extending the measurement capabilities of currently available sensors. The sensor consists of two electrodes embedded in a sand–plaster matrix that is stabilized with polyacrylamide. The sensor was calibrated and tested to determine its sensitivity to soil texture, temperature, and electrical conductivity. Experimental results showed that the sensor was able to determine matric potentials in the range of –7.5 kPa to –10 MPa, and showed excellent precision, with a RMSE <5 kPa across the range of –5 to –80 kPa. A standard electrical resistance–water potential curve can be established in the laboratory on a soil, and can then be applied to different soil types.
Abbreviations: PAM, polyacrylamide • TDR, time domain reflectometry
|
INTRODUCTION
|
|---|
The matric potential of water in soil, , is an important parameter for irrigation management and for quantifying unsaturated hydraulic properties of soil. The matric potential is often measured with water-filled hydraulic tensiometers; however, their measurement range is limited to > –80 kPa (Young and Sisson, 2002), and they usually require regular maintenance, although recent developments have allowed automatic refilling of tensiometers (Morrison and Szecsody, 1987; Faybishenko, 2000).
Another method to measure soil water potential is thermocouple psychrometry. Although thermocouple psychrometry can be used for a relatively wide range, the accuracy is better in drier conditions, and the upper measurement limit is about –30 to –200 kPa (Andraski and Scanlon, 2002). Furthermore, the measurements are very sensitive to temperature (Rawlins and Campbell, 1986), which limits the suitability of psychrometers for field applications.
Soil water potential can also be inferred by indirect methods, including heat dissipation, time-domain reflectometry (TDR), and electrical resistance measurements. The heat dissipation method is based on measuring the water-content-dependent heat dissipation in a porous matrix (Flint et al., 2002). Or and Wraith (1999) developed a sensor based on a coaxial cage embedded in a porous disk and measured electrical permittivity with TDR. Noborio et al. (1999) embedded a two-rod TDR waveguide in a gypsum block to determine soil water potentials. A similar design was used by Whalley et al. (2001), who embedded a three-rod waveguide in a ceramic matrix and measured the dielectric permittivity in the frequency domain. The soil water potential was then related to dielectric permittivity via a calibration equation.
Electrical resistance sensors operate with a similar principle as heat dissipation or electrical permittivity sensors by inferring the water potential from indirect mesurements. The water-content-dependent electrical resistance of a porous matrix embedded in soil is measured and related to matric potential (Scanlon et al., 2002). The porous matrix of the electrical resistance sensors is usually made of gypsum to control the salinity sensitivity of resistance measurements. The electrical resistance sensors are relatively inexpensive and have been widely used in irrigated agriculture (Thomson and Ross, 1996; Abraham et al., 2000; Miranda et al., 2005; Intrigliolo and Castel, 2006). The conventional gypsum electrical resistance block fails, however, when the water potential is greater than –30 kPa (Bourget et al., 1958). It was reported that the use of different porous materials allowed the measurement range to be expanded (Perrier and Marsh, 1958; Or and Wraith, 1999). For instance, the Watermark sensor (Campbell Scientific, Logan, UT) was optimized for use between –10 and –100 kPa (Spaans and Baker, 1992). Abraham et al. (2000) compared electrical resistance sensors made of different porous matrices, i.e., gypsum, soil, washed sand, sponge, and nylon. Based on reproducibility of measurements, they concluded that the washed sand provided the best matrix for the resistance measurements; However, they provided no details on how the sensors were constructed and how electrolyte concentrations in the sensor were controlled. Table 1 shows an overview of different water potential sensors and their measurement range.
It has been reported that certain water potential sensors have poor reproducibility. For instance, Spaans and Baker (1992) found that the Watermark sensor was not reproducible for a given soil. Whalley et al. (2001) concluded that poor reproducibility of electrical resistance sensors was often due to incomplete hydraulic contact and the mismatch of the air-entry potentials between sensor matrix and the surrounding soil.
The objective of this study was to design a sensor that could measure accurately and precisely across a wide range of moisture conditions. We developed an electrical resistance sensor that was constructed with a sandy soil plaster. We describe the sensor design, calibration, and its sensitivity to temperature, electrical conductivity, and soil texture.
|
MATERIALS AND METHODS
|
|---|
Sensor Design and Construction
A schematic diagram of our resistance matric potential sensor is shown in Fig. 1. Two parallel stainless steel electrodes (50 mm long and 0.4 mm in diameter) were inserted into a porous matrix, which was packed into a ceramic cup. The ceramic cup was usually used for water-filled hydraulic tensiometers and was purchased from a commercial supplier (Nanjing Ceramics Manufacturer, Nanjing, China). Its bubbling pressure was about 80 kPa. The porous matrix consisted of a mixture of gypsum powder, soil, and polyacrylamide (PAM). The soil samples were collected from agricultural fields near the Fengqiu Experimental Station of the Chinese Academy of Sciences, located in the Henan province of China. Soil organic matter was removed by adding 6% H2O2, and the soil was then oven dried at 110°C. The soil was mixed with different ratios of soil to gypsum (gypsum powder), and then passed through a 1-mm sieve. Polyacrylamide (1:1000 w/w polyacrylamide/soil–gypsum ratio) was added to provide structural stability to the soil–gypsum matrix. The PAM was purchased from the Shanghai Chemical Reagent Corp., Shanghai, China (Lot no. F20021129). The air-dry mixture was then packed into the ceramic cups and the electrodes were inserted into the packed material. The material was then saturated with deionized water and sealed at the top with epoxy glue. Finally, a polyvinyl chloride (PVC) tube was glued onto the ceramic cup.
The electrical conductivity between the two electrodes was measured with a SY-3 digital device for measuring electrical conductivity (Nanjing Quark Scientific Co., Nanjing, China). We calibrated the device with a series of standard resistances (50, 100, 500, 1000, 2000, 5000, 10000, 50000, and 100000 
). To prevent polarization of the electrodes, we used an AC current for the measurements.
Screening tests with different soil textures (sand and loam), different soil/gypsum ratios (4:1; 3:2; and 1:1 w/w), and different bulk density packings (1.36, 1.12, and 1.04 g cm–3) were conducted to select the most suitable combination for use in the resistance sensor. These screening tests were based on measurement reproducibility. The optimal combination was a sandy soil, a soil/gypsum ratio of 3:2 w/w, and a bulk density of 1.12 g cm–3. Figure 2 illustrates the effect of PAM on measurement reproducibility for two drying experiments, demonstrating that the polymer indeed considerably improved the sensor performance.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 2. Effect of polyacrylamide (PAM) on reproducibility of resistance measurements with the same sensor: (a) with PAM, (b) without PAM. Arrows indicate sequence of measurements (drying); the sensor was installed in the soil sample and water was removed by air drying.
|
|
Temperature and Electrical Conductivity Sensitivity
To test the temperature sensitivity of the resistance measurements, we installed two electrical resistance sensors and a temperature sensor (HEL-777 Pt resistance temperature sensor, Honeywell Sensing and Control, Golden Valley, MN) into a soil column of 15-cm diameter and 20-cm length. The soil in the column was the same as the one used for sensor construction. We wetted the soil to a water content of about 0.2 kg kg–1, and then sealed the column to prevent evaporation. The temperature was controlled by an environmental chamber in which the air temperature varied from 5 to 40°C in 5°C increments. For each temperature step, soil temperatures were allowed to equilibrate for >12 h before resistance measurements were taken.
To investigate the impact of electrical conductivity on measured resistance, we immersed three sensors in the following NaCl solutions: 0 (deionized water), 0.002, 0.005, 0.01, 0.02, 0.05, and 0.1 M. The electrical conductivity of each solution was measured using a conductivity electrode calibrated with a standard solution.
Soil Matric Potential Calibration
Each sensor was calibrated in the sandy soil, which was used for sensor construction. Table 2 lists the major properties of the soil. Particle-size analysis was performed using the pipette method (Gee and Bauder, 1986), and the soil organic C was measured using the Walkley–Black titration method (Nelson and Sommers, 1982). We determined the water retention curve for the soil by using hanging water columns, pressure plates, and vapor equilibration techniques (Table 3; Dane et al., 2002).
View this table:
[in this window]
[in a new window]
|
Table 3. Water retention functions for the sandy soil determined by hanging water columns, pressure plates, and vapor equilibration techniques.
|
|
For sensor calibration, the soil was first oven dried at 110°C for 1 d. Eight resistance sensors, two temperature sensors, and two hydraulic Hg tensiometers were installed into 5-cm depth of a dry soil column (20-cm i.d. and 10 cm in height). The resistance measurements were corrected for temperature fluctuations as described below.
The soil was first saturated with tap water (electrical conductivity = 0.297 dS m–1, total ionic concentration = 0.015%) from the bottom, and then dried slowly by continuous evaporation from the soil surface. Soil samples were taken from the 2- to 8-cm layer and the water contents were determined gravimetrically. For the water potentials outside the range of the hydraulic tensiometers, we used the previously measured water retention curve to infer water potentials from gravimetric water contents. Electrical resistance measurements were then related to the matric potential to obtain a calibration curve for each sensor.
It is known that gypsum block sensors do not operate reliably under wet conditions, i.e., when the water potential is more than –100 kPa. The reason is that the porous matrix made of Plaster of Paris has an air-entry potential less than –100 kPa and the matrix does not drain until the soil dries beyond the air-entry potential (Whalley et al., 2001). Our sensor was designed to operate at potentials greater than –100 kPa, and we extensively tested its operation in the range from 0 to –100 kPa in the following wetting–drying cycles.
The soil was subjected to three wetting–drying cycles after saturation with tap water from the bottom of the soil column. The first wetting–drying cycle was used to calibrate the electric resistance sensor based on the tensiometer measurements. The remaining wetting–drying cycles were used to check sensor repeatability (accuracy and precision). The sensor accuracy was assessed by the RMSE, and the precision was assessed by repeated measurements at a given water potential.
We further installed the electrical resistance sensors in a silty clay soil column (20-cm i.d. and 10 cm in height; Table 2) and then used the calibration obtained from the sandy soil to check whether a universal sensor calibration could be used to measure water potentials in different soils.
Response Time
To evaluate the response time of the electrical sensor, three sensors was saturated with tap water for 1 d, and were installed into a sandy soil column (10-cm i.d. and 10 cm in height) that was equilibrated on a pressure plate with a 100-cm water column for 5 d. Temporal changes in the sensor resistance were measured at 5-min to 12-h intervals.
|
RESULTS AND DISCUSSION
|
|---|
Temperature and Electrical Conductivity Sensitivity
Figure 3 shows the change in the electrical resistance readings in the sandy soil with varying temperature for two sensors. For each sensor, two measurements were taken at each temperature and the average was reported. The measured electrical resistance RT at a temperature T (°C) could be related to a standard temperature by (Spaans and Baker, 1992)
 | [1] |
where R25 is the measured resistance at 25°C, and the factor a is usually between 0.02 and 0.03 (Campbell and Gee, 1986; U.S. Geological Survey, 1998). For our sensors, the value of a was 0.0245, very close to the value of a = 0.024 reported by Spaans and Baker (1992).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 3. Measured resistance as a function of soil temperature for two sensors. Each data point represents an average of two measurements. RT/R25 is the ratio of the measured resistance at a temperature T (°C) and 25°C.
|
|
Figure 4 shows measured resistance of three sensors as a function of a series of NaCl solutions with different electrical conductivities. With an increase in electrical conductivity, measured resistance generally decreased. For all three sensors, increasing the solution electrical conductivity from 1.5 x 10–3 dS m–1 (deionized water) to 0.567 dS m–1 had no noticeable effect on the measured resistance, which was about the same as that of deionized water. When electrical conductivity was >1.12 dS m–1 (0.01 M), there was an obvious decrease in measured resistance for the three sensors. This shows that the resistance sensors cannot be applied to highly saline soils.
Soil Matric Potential Calibration
For a typical resistance sensor, a log–log plot of resistance vs. matric potential was generally linear (Bouyoucos and Mick, 1940), that is,
 | [2] |
where R is the resistance between the two electrodes (), is the soil matric potential (kPa), and b and c are constants. In our study, the relationship between electrical resistance and soil matric potential can be well described by Eq. [1]. For all eight calibrated sensors, we found that a linear calibration equation provided excellent fit to the experimental data with R2 ranging between 0.951 and 0.991 (Table 4). Different sensors had different slope and intercept parameters in the calibration equation. This shows that each sensor has to be calibrated separately. This finding is consistent with the calibration results reported by Spaans and Baker (1992) and Hymer et al. (2000).
The results also showed that the new sensors could be calibrated across a range from –7.5 kPa to about –10 MPa (Fig. 5). This is a much larger range than the one achievable with the commercial gypsum block sensors (Table 1).
Sensor Accuracy and Precision
Figure 6 shows the water potential data obtained from the three wetting–drying cycles for Sensor 1. The first cycle was used to calibrate the sensor and the subsequent cycles were used to assess sensor repeatability. Measurements with all eight sensors, with matric potentials ranging from –5 to –80 kPa, yielded a mean RMSE of 1.6 kPa for the first wetting–drying cycle (calibration), 1.86 kPa for the second cycle, and 3.45 kPa for the third cycle. In the third cycle, there were relatively large errors at potentials lower than –40 kPa, and we attributed this to hysteresis of the sensor matrix. Hysteresis was a problem for all the sensors when matric potential was measured in porous media. Nonetheless, the measurements showed that, across a range of –5 to –80 kPa, the sensor was accurate and reproducible.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 6. Comparison of soil matric potentials measured with resistance sensors vs. tensiometers in the sandy soil. The first wetting–drying cycle was used for sensor calibration and the subsequent cycles for sensor validation.
|
|
Application of a Universal Calibration Equation
Figure 7 shows that when the sensors were calibrated in the sandy soil, they could be used to determine the water potentials in a silty clay soil. Representative results for two sensors are shown in the figure. In the silty clay soil, the largest deviation between the resistance sensor and tensiometer data occurred between –40 and –70 kPa. The RMSE for all sensors are listed in Table 5. The mean RMSE in the silty clay soil was 4.6 kPa, compared with a RMSE of 2.6 kPa in the sandy soil. Sensor 8 showed large deviations between –40 and –70 kPa, resulting in a large RMSE (Table 5). We observed that cracks occurred at the soil surface during soil drying, which might have led to nonuniform water potentials in the soil. Overall, the data showed that the single calibration in the sandy soil was applicable to the silty clay soil.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 7. Soil matric potential determined by resistance sensors and tensiometers in two different soils. The sand was used for calibration and the silty clay soil was used for validation.
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Calculated RMSE of the electrical resistance sensors against matric potential measurements from tensiometers in two soils across the range of 0 to –80 kPa.
|
|
Response Time
Temporal changes in resistance for three sensors are shown in Fig. 8. After the sensors were installed into the soil column, the resistance of the sensors started to increase and reached the first equilibration in 12 h, and the values of measurements in the following 40 h were very stable. We therefore considered that about 12 h was required to reach equilibrium from saturation to = –10 kPa. Response time might be longer under other conditions, however, because it is related to soil texture, water content, and bulk density as well. In the experiments, we also measured the response time during the wetting processes from –80 kPa to saturation. It took the sensor 5 to 9 h to reach equilibrium. The response time may apply to field conditions where soil is dried by evapotranspiration; however, matric potential measurements under conditions of rapid change in water status may be less reliable.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 8. Response time of three electrical sensors during water desorption from saturation to a matric potential of –10 kPa.
|
|
|
CONCLUSIONS
|
|---|
The objective of this study was to develop a new electrical resistance sensor that could measure the matric potentials precisely across a wide range. The core of the sensor consisted of two electrodes embedded in a ceramic cup filled with sandy soil and gypsum powder. The porous medium was stabilized by adding PAM. The temperature-corrected resistance value was converted to soil matric potentials using a logarithmic calibration curve. The sensor could determine water potentials in the range from –7.5 kPa to –10 MPa. The sensor showed excellent precision and could be used in different soils when calibrated in one soil. Moreover, the sensor is inexpensive and could be automated and maintained readily. Because the sensor was sensitive to electrical conductivity, and the response time was relatively long, however, it cannot be applied to high-salinity soils or under conditions of rapid water content changes.
|
ACKNOWLEDGMENTS
|
|---|
We thank Dr. M. Flury for his constructive suggestions. This research was funded by the National Basic Research Program of China (Project no. 2005CB121103) and the Hi-Tech Research and Development Program of China (Project no. 2006AA10A301).
|
NOTES
|
|---|
A patent application was submitted for the sensor described in this study (China Patent Application no. 200510041392.7).
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication May 22, 2006.
|
REFERENCES
|
|---|
- Abraham, N., P.S. Hema, E.K. Saritha, and S. Subramannian. 2000. Irrigation automation based on soil electrical conductivity and leaf temperature. Agric. Water Manage. 45:145–157.[CrossRef]
- Andraski, B.J., and B.R. Scanlon. 2002. Thermocouple psychrometry. p. 609–642. In J.H. Dane and G.C. Topp (ed.) Methods of soil analysis. Part 4. Physical methods. SSSA Book Ser. 5. SSSA, Madison, WI.
- Bourget, S.J., D.E. Elrick, and C.B. Tanner. 1958. Electrical resistance units for moisture measurements: Their moisture hysteresis, uniformity and sensitivity. Soil Sci. 86:298–304.
- Bouyoucos, G.J., and A.H. Mick. 1940. An electrical resistance method for the continuous measurement of soil moisture under field conditions. Bull. 172. Michigan Agric. Exp. Stn., East Lansing.
- Campbell, G.S., and G.W. Gee. 1986. Water potential: Miscellaneous methods. P. 619–633. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. SSSA Book Ser. 5. SSSA, Madison, WI.
- Campbell Scientific. 1993. Model 257 and 257-L (Watermark 200) soil moisture sensor instruction manual. Campbell Scientific, Logan, UT.
- Dane, J.H., J.W. Hopmans, N. Romano, J. Nimmo, and K.A. Winfield. 2002. Soil water retention and storage—Laboratory methods. p. 675–720. In J.H. Dane and G.C. Topp (ed.) Methods of soil analysis. Part 4. Physical methods. SSSA Book Ser. 5. SSSA, Madison, WI.
- Delta-T Devices. 1999. Equitensiometer soil matric potential sensor type EQ2 user manual. Available at www.delta-t.co.uk/support-article.html?article=faq2005100703246 (verified 16 Feb. 2007). Delta-T Devices, Cambridge, UK.
- Faybishenko, B. 2000. Tensiometer for shallow and deep measurements of water pressure in vadose zone and groundwater. Soil Sci. 165:473–482.[CrossRef]
- Flint, A.L., G.S. Campbell, K.M. Ellett, and C. Calissendorff. 2002. Calibration and temperature correction of heat dissipation matric potential sensors. Soil Sci. Soc. Am. J. 66:1439–1445.[Abstract/Free Full Text]
- Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 383–411. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. SSSA Book Ser. 5. SSSA, Madison, WI.
- Hymer, D.C., M.S. Moran, and T.O. Keefer. 2000. Soil water evaluation using a hydrologic model and calibrated sensor network. Soil Sci. Soc. Am. J. 64:319–326.[Abstract/Free Full Text]
- Intrigliolo, D.S., and J.R. Castel. 2006. Performance of various water stress indicators for prediction of fruit size response to deficit irrigation in plum. Agric. Water Manage. 83:173–180.[CrossRef]
- Miranda, F.R., R.E. Yoder, J.B. Wilkerson, and L.O. Odhiambo. 2005. An autonomous controller for site-specific management of fixed irrigation systems. Comput. Electron. Agric. 48:183–197.[CrossRef]
- Morrison, R.D., and J.E. Szecsody. 1987. A tensiometer and pore water sampler for vadose zone monitoring. Soil Sci. 144:367–372.
- Nelson, D.W., and L.E. Sommers. 1982. Total carbon, organic carbon, and organic matter. p. 539–579. In A.L. Page (ed.) Methods of soil analysis. Part 2. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- Noborio, K., R. Horton, and C.S. Tan. 1999. Time domain reflectometry probe for simultaneous measurement of soil matric potential and water content. Soil Sci. Soc. Am. J. 63:1500–1505.[Abstract/Free Full Text]
- Or, D., and J.M. Wraith. 1999. A new soil matric potential sensor based on time domain reflectometry. Water Resour. Res. 35:3399–3407.[CrossRef]
- Perrier, E.R., and A.W. Marsh. 1958. Performance characteristics of various electrical resistance units and gypsum materials. Soil Sci. Soc. Am. Proc. 22:140–147.
- Rawlins, S.L., and G.S. Campbell. 1986. Water potential: Thermocouple psychrometry. p. 597–618. In A. Klute (ed.) Methods of soil analysis. Part 1. Agron. Monogr. 9. 2nd ed. ASA and SSSA, Madison, WI.
- Scanlon, B.R., B.J. Andraski, and J. Bilskie. 2002. Miscellaneous methods for measuring matric or water potential. p. 643–670. In J.H. Dane and G.C. Topp (ed.) Methods of soil analysis. Part 4. Physical methods. SSSA Book Ser. 5. SSSA, Madison, WI.
- Spaans, E.J.A., and J.M. Baker. 1992. Calibration of watermark soil moisture sensors for soil matric potential and temperature. Plant Soil 143:213–217.[CrossRef][Web of Science]
- Thomson, S.J., and B.B. Ross. 1996. Model-based irrigation management using a dynamic parameter adjustment method. Comput. Electron. Agric. 14:269–290.[CrossRef]
- U.S. Geological Survey. 1998. Specific electrical conductance. Tech. Water Resour. Invest. 9. USGS, Denver.
- Wescor, Inc. 1998. Water potential systems. Wescor, Logan, UT.
- Whalley, W.R., C.W. Watts, M.A. Hilhorst, N.R.A. Bird, J. Balendonck, and D.J. Longstaff. 2001. The design of porous material sensors to measure the matric potential of water in soil. Eur. J. Soil Sci. 52:511–519.[CrossRef]
- Young, M.H., and J.B. Sisson. 2002. Tensiometry. p. 575–609. In J.H. Dane and G.C. Topp (ed.) Methods of soil analysis. Part 4. Physical methods. SSSA Book Ser. 5. SSSA, Madison, WI.
TOP
ABSTRACT
INTRODUCTION
