HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Polyakov, V. Articles by Ryder, M. H. Search for Related Content PubMed Articles by Polyakov, V. Articles by Ryder, M. H. GeoRef GeoRef Citation Agricola Articles by Polyakov, V. Articles by Ryder, M. H. Related Collections Water Content Soil Methods/Instrumentation

SPECIAL SECTION: SOIL WATER SENSING

Calibration of a Capacitance System for Measuring Water Content of Tropical Soil

Department of Natural Resources and Environmental Management, University of Hawaii, 1910 East-West Road, Honolulu, HI 96822
^* Corresponding author (viktor{at}hawaii.edu)

Received 22 February 2005.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Capacitance sensors have improved substantially in the last decades, resulting in their wide acceptance. A new generation of multisensor capacitance systems (MCS) is now available that are easy to install and use. Calibration of capacitance sensors was conducted for a weathered clay loam soil and silica sand in field and laboratory conditions. The specific objectives of this research were to (i) conduct field and laboratory calibration of a new MCS in silica sand and soil, (ii) evaluate the performance of MCS for a shrinking–swelling tropical soil, and (iii) evaluate the effect of medium temperature on the MCS reading at constant water content. Three-parameter power type calibration equations were developed. The laboratory column calibration had higher correlation coefficients (R² = 0.96 and 0.97 for soil and sand, respectively) than the rangeland (R² = 0.73) and cultivated soils (R² = 0.74). The manufacturer default model fitted the field data reasonably well in the higher moisture range (0.35–0.45 cm³ cm^–3). However, it performed poorly in the dryer range (0.2–0.35 cm³ cm^–3), severely underestimating soil moisture content. Shrinking and swelling of soil and the presence of bound water might have affected the sensor's performance. Across the 45°C interval, there was 15% overestimation of the actual water content for soil and only 10% for sand. The relationship was statistically highly significant (P < 0.001) with an R² = 0.99 for both sand and soil. Use of MCS is suitable for tropical soil; however, site specific calibration is needed to improve the estimates of soil water content.

Abbreviations: MCS, multisensor capacitance system • TDR, time domain reflectometry

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
SOIL WATER CONTENT dynamics are closely related to soil physical properties, climate, and land management. The knowledge of soil water is critical for effective environmental protection and optimum agricultural production. Maintaining optimal soil water content is difficult without real-time soil water monitoring. Hence, there is a need for instrumentation that provides accurate real-time water content measurement at reasonable cost. Considering the spatial variability of soil physical properties, this instrumentation should allow multi-site and multi-depth measurements with minimal soil disturbance.

Capacitance soil water sensors operate at a narrow band frequency and use dielectric constant (K) of soil–water–air mixture to estimate soil water content. Water molecules, being permanent dipoles, respond to the electrical field by becoming polarized. The K of water (78.54 at 22°C) is large compared with those of soil matrix (<10) and air (1); thus, a change in soil water content will strongly influence the K of soil–water–air mixture. However, great variability of the K of soil minerals (4–9) (Robinson, 2004) and dry plant tissue (1–4) makes it necessary to calibrate these sensors for a particular soil (Baumhardt et al., 2000) and, if practical, for each soil horizon. In addition, K is a function of the ratio of free water to that of bound water, soil temperature, bulk density, and water salinity, especially at low sensor frequencies (Paltineanu and Starr, 1997). In most cases, the relationship between the MCS output and volumetric soil water content (_v) is a three-parameter power function (Fares et al., 2004; Paltineanu and Starr, 1997). However, linear approximation was also used for some soils (Gaudu et al., 1993). The nonlinearity of the relationship between MCS output and _v is partially attributed to clay-bound water, which has a behavior that is different from that of free water under the influence of electromagnetic waves (Bridge et al., 1996; Wang, 1980).

The effect of soil temperature on capacitance and time domain reflectometry (TDR) sensors has been reported for different soil types (Baumhardt et al., 2000). In the 5 to 45°C range, the K of water decreases with the increase of temperature at an approximate rate of 0.36°C^–1 (Weast, 1986). In addition, a small decrease of capacitance frequency, which means an apparent increase in water content, may be attributed to the temperature effects on the sensor's circuitry (Dean et al., 1987). Seyfried and Murdock (2001) reported negative effect of temperature on measured value of _v for sand, while for various soils the relationship was positive, resulting in large apparent _v variation across a 40°C temperature change. Kurà (1982), studying temperature effects on fine sand with a portable capacitance sensor at different water contents, reported an increase in the frequency of the capacitance sensor with increasing sand temperature from 15 to 30°C. The observed trend was much more evident at higher water contents. Baumhardt et al. (2000) observed cyclical fluctuation in _v induced by similar soil temperature fluctuations of 15°C as estimated by MCS and TDR. Wraith and Or (1999) demonstrated experimentally that temperature affects soil permittivity by changing the physical properties of water bound near the surface of 2:1 clay minerals. Furthermore, Yu et al. (1999) established theoretically that temperature changes in the mineral soil fraction have little impact on soil permittivity. Baumhardt et al. (2000) attributed _v fluctuations indicated by both TDR and MCS to temperature dependent fluctuations in the soil permittivity.

Recent advances in microelectronics have made capacitance sensors popular in situ soil water content monitoring devices (Fares and Alva, 2000). Capacitance sensors have been used to measure soil water content in a wide range of soil types by researchers and growers in Australia (Buss, 1993; Fares et al., 2004) and the USA (Baumhardt et al., 2000; Fares and Alva, 2000), for various applications such as irrigation scheduling (Girona et al., 2002) and waste water treatment. Capacitance sensors have been calibrated in the field (Fares et al., 2004; Morgan et al., 1999) and laboratory (Baumhardt et al., 2000; Paltineanu and Starr, 1997) conditions; however, limited number of calibration exercises were conducted using weathered tropical soils where bound water effect is present (Wu, 1998).

The objectives of this research were to (i) conduct field and laboratory calibration of a new MCS in silica sand and soil, (ii) evaluate the performance of MCS for a shrinking–swelling tropical soil, and (iii) evaluate the effect of medium temperature on the MCS reading at constant water content.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Capacitance Sensors
A multisensor capacitance system, EasyAg 50 (Sentek, Pty., Ltd., Stepny, South Australia), was used in both field and laboratory calibration experiments. The principles of operation and design of the MCS were described in detail by Buss (1993), Paltineanu and Starr, (1997) and Fares and Alva (2000). The system consists of multisensor probes and a data logger capable of operating 32 sensors (8 probes with 4 sensors each) simultaneously. An individual probe consists of an access tube, printed circuit board, and four capacitance sensors. Access tube is a 60-cm-long special PVC pipe with external and internal diameters of 32 and 28 mm, respectively. The diameter of the capacitance ring is such that it allows the sensor to move freely inside the access tube, but with minimum air gap between the rings and access tube wall. Data are downloaded from the data logger using specialized software that allows setting and modifying logger configuration, viewing, and exporting data.

Each capacitor sensor consists of two metal rings mounted on the circuit board at distances of 10, 20, 30, and 50 cm from the top of the access tube. These rings are a pair of electrodes that form the plates of the capacitor, with the soil acting as the dielectric between. The plates are connected to an oscillator, consisting of an inductor and a capacitor. The oscillating electrical field is generated between the two rings and extends into the soil medium through the wall of the access tube (99% of the reading is taken within a 10-cm radius around the sensor axis). The capacitor and the oscillator form a circuit, and changes in dielectric constant of surrounding media are detected by changes in the operating frequency. The capacitance sensors are designed to oscillate in excess of 100 MHz inside the access tube in free air. The output of the sensor is the frequency response of the soil's capacitance due to its soil moisture level. The resonant frequency of oscillation F is related to the capacitance as a function (Paltineanu and Starr, 1997):

[1]

where C is the total capacitance and L is the inductance of the circuit. The surrounding soil with all its components and access tube wall contribute to the capacitance C. The sensors were normalized by placing each access tube into a water bath and in the air at 22°C. Respective frequency readings for each sensor in both media were recorded. These readings are used to calculate the scaled frequency (SF) each time the sensor takes a measurement, using the following equation:

[2]

where SF is the scaled frequency, and F_a, F_s, and F_w are frequency readings of the sensor in air, soil, and water, respectively. The SF is then converted to _v using a default or user-specified calibration equation. A default calibration equation is provided by the manufacturer and is based on calibration in a variety of soils. In this study, calibration of capacitance sensors was conducted in four media: rangeland soil profile, disturbed soil profile on a cultivated terrace, laboratory soil, and sand columns with average bulk densities of 1.25, 1.17, 1.24, and 1.43 g cm^–3, respectively.

Experimental Site and Soil Properties
Field calibration of the capacitance sensors was conducted at the University of Hawaii-Manoa Poamoho Agricultural Research Station, Oahu, Hawaii (21.55° N; 158.12° W). The long-term average annual precipitation at the site is 1132 mm yr^–1, most of which occurs between November and April (USDA, 1972). The mean annual soil temperature is 23°C. Soil is classified as Ewa silty clay loam (fine, kaolinitic, isohyperthermic Aridic Haplustolls) and is found in basins and on alluvial fans (primary agricultural areas) on the islands of Maui and Oahu. It was formed on alluvium derived from igneous rock. The soil is moderately permeable, sticky, and plastic with granular structure and many fine and very fine tubular pores. The shrink–swell potential is moderate. Selected physical properties of the soil are presented in Table 1. The surface layer of the soil is dark reddish brown and is 45 cm thick. The subsoil consists of 1-m-thick dark reddish-brown and dark-red silty clay loam with subangular blocky structure. The substratum is gravelly alluvium (USDA, 1972).

Calibration Procedure
Six probes were installed on the rangeland site and two probes on the cultivated terrace site following the procedure recommended by the manufacturer (Sentek, 2003). To cover a wide range of soil water content, three moisture levels were used: dry, moist, and wet. For the wet treatment, for 2 d before sampling water was applied to the soil surface around the access tubes using specially designed drippers. The sampling was conducted shortly after the end of the wetting. For the moist treatment, water was applied in a similar way, and sampling performed 1 d after the end of the wetting period. For the dry treatment, no water was added. When soil was sampled at all four depths near the probe, the probe was relocated to a new location following the same installation procedure.

The soil samples were collected using an intact core sampler with internal diameter of 49 mm equipped with a drop hammer. The sampler had a brass sampling ring 76 mm in length. Intact soil cores were extracted in triplicate in close proximity to the access tube at every sensor depth (10, 20, 30, and 50 cm), such that the middle of the soil core was at the same depth as the center of individual sensor in the access tube. Soil samples were wrapped in plastic and transported to the laboratory, where they were weighed, oven dried at 105°C for 72 h, and reweighed to determine the bulk density and water content. Sensors were logged automatically at 1-min intervals. At least 10 readings preceding the sampling were used to obtain the sensor's average reading.

The laboratory calibration was conducted in a plastic cylinder (60 cm high and 20 cm in diameter). Starting with an air-dried soil an incremental amount of water (250 mL) was added and the soil was carefully mixed. The mixture was passed through the sieve and clods were carefully broken. The soil was placed into the column with sensor access tube in the middle and compacted to the bulk density of 1.25 g cm^–3. To ensure uniform bulk density throughout the soil column, the soil was added and compacted in small increments. The top of the soil column was tightly covered with plastic wrap to prevent water losses from the surface. The soil column with capacitance probe inserted in it was left to equilibrate for at least 24 h. Soil moisture readings were logged every minute. Ten consecutive readings collected at the end of the equilibration period were used to calculate average sensor readings. Soil samples were also collected near the sensors to determine actual water content gravimetrically and bulk density. Next, all the soil from the cylinder was remixed with additional soil from the original supply to higher water content level. The same process, described above, was repeated nine times to cover the water content range between air dry and saturated water content. A similar procedure was used with silica sand.

Multisensor capacitance system response to temperature of the media was evaluated using four-sensor probes in soil and sand at water contents 0.26 and 0.09 cm³ cm^–3, respectively. Soil and sand were placed in two separate columns with access tubes as described above and sealed to avoid loss of water. The columns were placed in an oven at 50°C until their internal temperature equilibrated with the oven temperature; they were then removed from the oven and allowed to cool gradually. Soil and sand inside the columns reached room temperature (22°C) over 12-h period. The temperatures and capacitance sensor readings were logged automatically at 1-min interval. The temperature of the media was monitored using copper-constantan thermocouples inserted into the middle of each of the columns. After reaching room temperature, both soil and sand columns were transferred into a cooling chamber and cooled down to 6°C in a 12-h period. Temperature, time, and sensor readings were logged as previously described.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Effect of Soil Depth and Water Content on Bulk Density
Experimental data show that soil water content measurement by the capacitance sensors is strongly influenced by soil bulk density (Huang et al., 2004). A comparison of measurements made in several soils by Gardner at el. (1998) revealed that differences in dry bulk density had greater effect on sensor reading than clay and organic matter content. Using five different sensors Huang at el. (2004) showed that the higher the soil bulk density, the greater the overestimation of soil water content.

There was no significant effect of soil water content on bulk density on the rangeland. Cracking of the soil during dry periods was observed on the rangeland; however, it was not reflected in overall bulk density, due to good soil structure. On the cultivated terrace there was highly significant positive correlation between water content and bulk density for all depths. A similar trend was observed by Fares et al. (2004) in sandy clay loam soil horizons. The effect of soil moisture content on bulk density was greater at deeper horizons (50 cm) than at those on the surface (10 cm) as indicated by the slope of the regression model, 1.49 and 0.69 for the 50- and 10-cm depths, respectively (Table 2).

[3]

where a, b, and c are fitting coefficients.

Separate calibration equations were developed for several media with increasing degree of dielectric complexity: sand column, soil column, cultivated site, and rangeland site. Laboratory calibration was conducted for wider range of _v (0.01–0.55 cm³ cm^–3) than that for the field condition (0.20–0.50 cm³ cm^–3) (Table 4, Fig. 1–4) . The laboratory column calibration had higher correlation coefficients (R² = 97 and 96% for sand and soil, respectively) than the cultivated (R² = 74%) and rangeland (R² = 73%) soils. Site-specific field calibrations had higher correlation coefficients than a calibration equation that used the pooled data from both sites (R² = 0.69).

View larger version (15K): [in this window] [in a new window]   Fig. 1. Measured volumetric water content versus scaled frequency and corresponding calibration models for Ewa silty clay loam on a rangeland site.

View larger version (18K): [in this window] [in a new window]   Fig. 4. Measured volumetric water content versus scaled frequency and corresponding calibration models for laboratory soil and sand columns.

View larger version (15K): [in this window] [in a new window]   Fig. 2. Measured volumetric water content versus scaled frequency and corresponding calibration models for Ewa silty clay loam on a cultivated terrace site.

View larger version (15K): [in this window] [in a new window]   Fig. 3. Measured volumetric water content versus scaled frequency and corresponding calibration models for all field data for the Ewa silty clay loam.

Differences between estimates of _v made using the default calibration and the measured _v for laboratory investigation were more pronounced and not as consistent for the Ewa soil than those for sand (Fig. 4 and 5) . The default calibration underestimated the water content for _v 0.2 cm³ cm^–3 and overestimated the water content for _v 0.2 cm³ cm^–3. Geesing et al. (2004) reported that for silty-loam soil, default calibration of their tested MCS generally underestimated the actual soil water content. This is not consistent with the results obtained by Baumhardt et al. (2000), who studied two fine-textured soils and reported that their calibration results significantly improved the accuracy of the _v estimation by the MCS in the 0.20 to 0.45 cm³ cm^–3 range as compared with either the manufacturer's or Paltineanu and Starr's (1997) calibration equations.

View larger version (21K): [in this window] [in a new window]   Fig. 5. Comparison of calibration models developed for rangeland, cultivated terrace, combined field data, laboratory soil and sand columns, and default manufacturer equation.

The field calibration data were more scattered and had larger RMSE than those of the sand and the laboratory Ewa soil. In addition, its water content–SF relationships was statistically less significant that those of the laboratory. Field calibration studies (Fares et al., 2004; Morgan et al., 1999) using either three or two-parameter power calibration model reported larger correlation coefficients than the current results (Table 4). Differences could be attributed to one or more of the following causes. First, the current study used a tropical soil that is different from either the sandy soil used by Morgan et al. (1999) or the sandy clay loam duplex soil used by Fares et al. (2004). Shrinking and swelling potential might have affected contact between access tube and the soil, which was reflected on the coefficient of correlation of calibration equation. Second, the MCS used in this study had smaller diameter than that used by Fares et al. (2004) or Morgan et al. (1999). This might have resulted in different zones of influence of the two systems, where the sensor with the smaller zone of influence is less accurate.

The manufacturer default model fitted the field data reasonably well (Fig. 1 and 2) in the wetter range (0.35–0.45 cm³ cm^–3, which corresponds to a SF between 0.86 and 0.95). However, it performed poorly in the dryer range (0.2–0.35 cm³ cm^–3), severely underestimating _v. The precision of the custom calibration equation in the field condition was ±0.08 cm³ cm^–3 in the mid section of the calibration curve. If the RMSE is considered an indicator of the accuracy of a regression model, RMSE values of the current field calibration are higher than those reported for MCS by Baumhardt et al. (2000) and Fares et al. (2004).

It has been reported (Dobson et al., 1985; Wang and Schmugge, 1980) that soil type affects the apparent K of soil, and therefore _v measurements based on K. The primary cause of difference between soils is usually attributed to the effects of solid–liquid interactions at the solid surface that restrict the rotational freedom of adsorbed water molecules (Seyfried and Murdock, 2001). This water, called bound water, is considered to have K much lower than that of the free water (Dobson et al., 1985), and its amount is expected to positively correlate with the soil surface area. Consequently, with the increase of clay content in the soil, the proportion of bound to free water will also increase, increasing the measured SF. Hence, _v is underestimated by the default equation at low moisture range, where the impact of bound water is more profound. This reasoning is supported by field data only (Fig. 1–3); however, the reverse is what was observed with the laboratory data, indicating that the wetting process may not reproduce the same moisture condition observed in the field as a result of the drying cycle.

Paltineanu and Starr (1997) argued that the most accurate calibration is achieved from carefully conducted laboratory calibration. Indeed, findings of our study for the soil column support this, as is shown by the higher R² (0.96 against 0.69) and lower RSME (0.038 against 0.046) than that obtained from the field data. In addition, working with sand was easier than working with soil, even in the laboratory. Sand was easier to mix and homogenize at low water content than Ewa clay loam. Ewa silty clay loam could also be more prone to localized excessive compaction and/or existence of air gaps than sand. These effects might cause the high variability encountered in Ewa clay loam calibration.

Temperature Effects
The relative response of the MCS to the temperature of the medium is presented in Fig. 6 . This figure shows the effect of temperature, ranging between 5 and 50°C, on the ratio of _T/_To, where _To is the actual water content at 23°C (average annual soil temperature) and _T is the apparent water content at a given temperature. The following exponential model was used to describe the temperature effect:

[4]

View larger version (16K): [in this window] [in a new window]   Fig. 6. The effect of temperature on volumetric water content measurement using MCS. Capacitance sensor measurement is represented as a ratio T/23, where T and 23 are the soil water content at temperature T and 23°C, respectively.

To correct for the temperature effects on _v, it is essential to account for the temperature effects on the sensor itself. The sensor used in the study is designed to oscillate at frequencies >100 MHz, which should minimize its sensitivity to the change in temperature. Dean et al. (1987) reported that for a high-frequency transistor used in MCS and operating at 150 MHz the frequency drift due to temperature is 0.5 MHz C°^–1. A correction for temperature effects on MCS water content estimate could be implemented within a calibration relationship, if soil temperature is measured along with water content (Baumhardt et al., 2000). This may be difficult to accomplish practically for several reasons. First, temperature effect is specific to the media. In this study the rate of _v measurement change with temperature in soil (0.001 cm³ cm^–3 °C^–1) was three times higher than that in sand (0.0003 cm³ cm^–3 °C^–1). Second, even for the same type of media (i.e., sand), some researchers reported negative temperature effects (Seyfried and Murdock, 2001), yet our study demonstrated positive effect. In addition, the temperature effect may vary depending on actual _v even within the same soil type (Baumhardt et al., 2000; Seyfried and Murdock, 2001). Ideally, an integrated temperature and capacitance sensor would be able to detect temperature changes in the surrounding soil and achieve dependable _v measurement.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Calibration of capacitance sensors was conducted for Ewa silty clay loam and silica sand in field and laboratory conditions. Field calibration was performed on rangeland and cultivated sites. The effect of soil temperature on capacitance sensors was examined. A wide range of soil water content (0.01–0.55 cm³ cm^–3) was used in the study. The depth of the soil profile had significant effect on soil bulk density in both rangeland and cultivated sites. In addition, soil bulk density on the cultivated site showed greater coefficient of variation (20%) than the rangeland (8%), indicating greater shrink–swell potential of undisturbed soil. The three-parameter power model described the relationship between the soil water content and the SF as measured by the MCS. The default model provided by the manufacturer fitted the field data well only in the higher water content range (0.35–0.45 cm³ cm^–3). It performed poorly in the lower range, severely underestimating the actual water contents. The precision of the custom calibration equation in the field condition was ±0.08 cm³ cm^–3 at = 0.05. These factors indicate the need for site-specific calibration if the MCS is to be practically used. The soil column calibration for both soil and sand had larger correlation coefficients (R² = 96 and 97%, respectively) than the rangeland (R² = 73%) and cultivated site (R² = 74%). The laboratory calibration data from the sand and soil were less scattered and had smaller RMSEs than the field data. The differences between the default calibration curve and the curve proposed for sand were small for most of the _v range (0–0.3 cm³ cm^–3); however, the default manufacturer calibration equation underestimated the _v in the sand in the upper range (0.3–0.5 cm³ cm^–3). These differences were attributed to the presence of bound water on the clay surfaces, which affected bulk permittivity, and shrinking and swelling of soil, which might have affected the contact between MCS probe and the media. The temperature of the media resulted in a 15% overestimation of the actual water content for Ewa silty clay loam and 10% for silica sand for the 45°C interval. This study demonstrated that site-specific calibration improved the accuracy of water content measurement by MCS as compared with the default equation. In addition, we quantified the effect of temperature on MCS reading and demonstrated the need to account for temperature variations by integrating capacitance and temperature sensors.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS 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]
• Bridge, B.J., J. Sabburg, K.O. Habash, J.A.R. Ball, and N.H. Hancock. 1996. The dielectric behaviour of clay soils and its application to time domain reflectometry. Aust. J. Soil Res. 34:825–835.[CrossRef]
• Buss, P. 1993. The use of capacitance based measurements of real time soil water profile dynamics for irrigation scheduling. Proc. Natl. Conf. Irrig. Assoc., Australia and Natl. Committee Irrig. Drain., Lauceston, Tasmania. 17–19 May 1993. Irrig. Assoc. of Aust., Homebush, NSW, Australia.
• Dean, T.J., J.P. Bell, and A.J.B. Baty. 1987. Soil moisture measurement by an improved capacitance technique. 1. Sensor design and performance. J. Hydrol. (Amsterdam) 93:67–78.
• Dobson, M.C., F.T. Ulaby, M.T. Hallikainen, and M.A. Elrayes. 1985. Microwave dielectric behavior of wet soil. 2. Dielectric mixing models. IEEE Trans. Geosci. Remote Sens. 23:35–46.
• Fares, A., and A.K. Alva. 2000. Evaluation of capacitance probes for optimal irrigation of citrus through soil moisture monitoring in an Entisol profile. Irrig. Sci. 19:57–64.[CrossRef]
• Fares, A., P. Buss, M. Dalton, A.I. El-Kadi, and L.R. Parsons. 2004. Dual field calibration of capacitance and neutron soil water sensors in a shrinking-swelling clay soil. Available at www.vadosezonejournal.org. Vadose Zone J. 3:1390–1399.[Abstract/Free Full Text]
• Gardner, C.M.K., T.J. Dean, and J.D. Cooper. 1998. Soil water content measurement with a high-frequency capacitance sensor. J. Agric. Eng. Res. 71:395–403.[CrossRef]
• Gaudu, J.C., J.M. Mathieu, J.C. Fumanal, L. Bruckler, A. Chanzy, P. Bertuzzi, P. Stengel, and R. Guennelon. 1993. Soil moisture measurement using a capacitive probe. Agronomie 13:57–73.
• Geesing, D., M. Bachmaier, and U. Schmidhalter. 2004. Field calibration of a capacitance soil water probe in heterogeneous fields. Aust. J. Soil Res. 42:289–299.[CrossRef]
• Girona, J., M. Mata, E. Fereres, D.A. Goldhamer, and M. Cohen. 2002. Evapotranspiration and soil water dynamics of peach trees under water deficits. Agric. Water Manage. 54:107–122.[CrossRef]
• Huang, Q., O.O. Akinremi, R.S. Rajan, and R. Bullock. 2004. Laboratory and field evaluation of five soil water sensors. Can. J. Soil Sci. 84:431–438.
• Kurà, V. 1982. Testing of a field dielectric soil moisture meter. ASTM Geotech. Test. J. 4:111–116.
• Morgan, K.T., L.R. Parsons, T.A. Wheaton, D.J. Pitts, and T.A. Obreza. 1999. Field calibration of a capacitance water content probe in fine sand soils. Soil Sci. Soc. Am. J. 63:987–989.[Abstract/Free Full Text]
• Paltineanu, I.C., and J.L. Starr. 1997. Real-time soil water dynamics using multisensor capacitance probes: Laboratory calibration. Soil Sci. Soc. Am. J. 61:1576–1585.[Abstract/Free Full Text]
• Robinson, D.A. 2004. Calculation of the dielectric properties of temperate and tropical soil minerals from ion polarizabilities using the Clausius-Mosotti equation. Soil Sci. Soc. Am. J. 68:1780–1785.[Abstract/Free Full Text]
• Sentek. 2003. EasyAG installation guide. Sentek Sensor Technologies, Stepney, Australia.
• Seyfried, M.S., and M.D. Murdock. 2001. Response of a new soil water sensor to variable soil, water content, and temperature. Soil Sci. Soc. Am. J. 65:28–34.[Abstract/Free Full Text]
• Seyfried, M.S., and M.D. Murdock. 2004. Measurement of soil water content with a 50-MHz soil dielectric sensor. Soil Sci. Soc. Am. J. 68:394–403.[Abstract/Free Full Text]
• USDA. 1972. Soil survey of the islands of Kauai, Oahu, Maui, Molokai, and Lanai, State of Hawaii. USDA-SCS, Washington, DC.
• Wang, J.R. 1980. The dielectric-properties of soil-water mixtures at microwave-frequencies. Radio Sci. 15:977–985.
• Wang, J.R., and T.J. Schmugge. 1980. An empirical model for the complex dielectric permittivity of soils as a function of water content. IEEE Trans. Geosci. Remote Sens. 18:288–295.
• Weast, R.C. (ed.) 1986. Handbook of chemistry and physics. CRC Press, Boca Raton, FL.
• 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.[CrossRef]
• Wu, K.G. 1998. Measurement of soil moisture change in spatially heterogeneous weathered soils using a capacitance probe. Hydrol. Processes 12:135–146.[CrossRef]
• Yu, C., A.W. Warrick, and M.H. Conklin. 1999. Derived functions of time domain reflectometry for soil moisture measurement. Water Resour. Res. 35:1789–1796.[CrossRef]

This article has been cited by other articles:

				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]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Polyakov, V. Articles by Ryder, M. H. Search for Related Content PubMed Articles by Polyakov, V. Articles by Ryder, M. H. GeoRef GeoRef Citation Agricola Articles by Polyakov, V. Articles by Ryder, M. H. Related Collections Water Content Soil Methods/Instrumentation