Soil Science Society of America Journal 63:1505-1512 (1999)
© 1999 Soil Science Society of America
DIVISION S-1-SOIL PHYSICS
Small-Scale Measurement of Soil Water Content Using a Fiber Optic Sensor
Fernando Garridoa,
Masoud Ghodratia and
Michael Chendoraina
a Division of Ecosystem Sciences, Dep. of Environmental Science, Policy and Management, Univ. of California, Berkeley, CA 94720-3110 USA
ghodrati{at}nature.berkeley.edu
 |
ABSTRACT
|
|---|
Many water flow and solute transport studies require accurate measurement of water content within small soil volumes. We have examined the feasibility of using fiber optic mini-probes (FOMPs) for in situ measurement of water content in relatively small soil volumes (<15–20 mm3) in real time and on a continuous basis. The system consists of transmitting a constant beam of light through the input leg of a fiber optic mini-probe to a location of interest within the soil matrix. At the tip, the light exits the probe, interacts with the soil volume directly in front, and partially reflects back into the probe. The reflected signal is transmitted through the output leg to a photodetector and quantified. The output signal, which is constant during steady state (i.e., dry soil), decreases as the water content in soil increases. A calibration is necessary to convert the output light intensity to water content. In developing calibration curves for the three soils used in the study, we consistently found an excellent correlation (r2 > 0.98) between the soil light reflectivity and the water content. Calibration of the FOMP depends on the individual probe, soil type (largely due to texture), and bulk density. The FOMP system may be ideal in situations where water content is dynamic and changing at small spatial scales, especially where these changes have a large impact on other processes.
Abbreviations: CV, coefficient of variation • FOMP, fiber optic mini-probe • TDR, time domain reflectometry
|
INTRODUCTION
|
|---|
MEASUREMENT OF ABSOLUTE SOIL WATER CONTENT, as well as monitoring relative changes in soil water content are among the most important physical measurements in soil (Nielsen et al., 1996). Aside from the routine measurements of soil water content made in agricultural and environmental settings, many water flow and solute transport studies require accurate knowledge of water content in some finite volume of soil. Depending on the nature of these studies, the measurement needs may extend anywhere from an estimation of the volume-averaged water content of an entire field (Rudolph et al., 1996) to very precise measurement of water content in a thin soil layer (Amato and Ritchie, 1995). In some situations, a simple measurement technique that only provides a "snap shot" of the water content may be sufficient, yet in others we may need methods that are nondestructive and make measurements of soil water content on a continuous basis. Since it is unlikely for a single methodology or technique to provide precise measurement across the entire required temporal and spatial scales, we need to develop specific methodologies for different scales.
Presently, a number of useful techniques are available for in situ measurement of water content in soil (for a good review of the topic see Topp et al., 1996). All of these methods, such as time domain reflectometry (TDR), heat-pulse sensors, neutron probe, portable dielectric probe, and capacitance measurement techniques, provide an averaged measurement of water content for a soil volume whose geometry cannot be exactly defined. All of these devices have significant advantages and disadvantages. For example, only recently specialized TDR coil probes have been developed to measure water content at smaller scales (Nissen et al., 1998). Other methods presently do not allow for measurement of water content below the cubic centimeter scale (Amato and Ritchie, 1995; Reece, 1996). Measurement of changes in water content below the centimeter scale can be quite useful, especially in studies investigating rhizosphere dynamics, soil microbial activity, and macropore flow mechanisms. Furthermore, many relevant processes at these scales require a high degree of temporal resolution, such as the formation of film flow to induce macropore flow (Tokunaga and Wan, 1997).
Fiber optic sensors, which are capable of providing continuous real-time measurements, are presently used in many fields of science to measure and monitor various physical and chemical processes. Since these sensors can be relatively small and can access a measurement arena with little disturbance, they may be especially suitable for the measurement of small-scale processes in soil, such as water dynamics.
|
Fiber Optic Sensors
|
|---|
Operation of most optical sensors is based on the measurement of changes that a process imposes on a constant beam of light (i.e., the quantification of differences between the input and the output light signal) (Krohn, 1988). Whereas in conventional optical systems the sample or the process of interest is brought to the measurement device (e.g., a cuvette inside a spectrometer), with fiber optic systems, light signals are carried to and from a remote measurement arena via small optical fibers. This feature allows for real time in situ measurements of various processes in a wide array of media and measurement environments on a continuous basis (Goyet et al., 1992; Maher and Shahriari, 1993; Consentino et al, 1995; Hadjiloucas et al., 1995; Rogers and Poziomek, 1996).
Since incident light is absorbed by water, a wetter soil is "darker" than a drier soil and reflects less light. This soil property can be exploited using a fiber optic sensor which measures the reflected light intensity at some remote point inside the soil matrix. For instance, the common end of a bifurcated (i.e., "Y"-shaped) fiber optic bundle that is encased in a small diameter stainless steel tubing and inserted into soil to a desired depth can serve as a simple FOMP or sensor (Ghodrati, 1999). In this simple design, a constant beam of light is sent through the input leg to the common end, where it exits the fiber optic bundle, interacts with the soil volume immediately in front of the probe and is partially reflected back into the probe. The reflected light is carried through the output leg to a photodetector and quantified. If the changes in the output light intensity are significantly correlated with the water content of the soil in front of the probe, then the device has the potential to serve as a moisture sensor.
Previous attempts to develop fiber optic soil moisture sensors have been based on the measurement of changes in the total internal reflection of light in chemically doped (Muto et al., 1990) or bent (Alessi and Prunty, 1986) fiber optic light guides. Although the principle is sound, these types of sensors are more difficult to develop, in most cases require installation in soil (rather than simple insertion), and as pointed out by Alessi and Prunty (1986), these sensors may not operate properly if installed in direct contact with soil and may need to be first embedded in some porous material.
The purpose of this study was to examine the feasibility of using a FOMP to measure in situ soil water content at small scales. With respect to the system's feasibility, we attempted to (i) quantify variability in terms of probe installation, soil texture, and soil reflectivity limits in dry and saturated conditions; (ii) quantify the relationship between soil reflectivity and water content measured with the fiber optics; and (iii) verify this relationship with independent measurements.
|
Materials and methods
|
|---|
Fiber Optic Mini-Probe (FOMP) System
The FOMP system used in this study is similar to the one described in detail in Ghodrati (1999). Briefly, the system consists of a (i) light source, (ii) fiber optic mini-probes (Fig. 1a)
consisting of bifurcated fiber optic bundles with diameters of 3.4 and 2.5 mm at the common end, (iii) a cadmium sulfide photoresistor-based photodetector, (iv) a multimeter, and (v) a computer as a data acquisition system. Filter combinations for light output and input can be used which specifically target water, thus increasing the sensitivity of the system.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1 Design of the fiber optic mini-probe (FOMP) system for measuring moisture content. (a) Schematic of a bifurcated fiber optic miniprobe. (b) Diagram of the FOMP soil moisture calibration chamber. (c) Diagram of the soil column used to evaluate the FOMP soil moisture calibration
|
|
An important question concerning the fiber optic mini-probe is the size and shape of the surface and volume of the soil measured. Very little experimentation has been performed on this topic. One study suggested that the fiber optic light penetration decreases exponentially up to a depth of 3 mm into the soil from the probe tip (Kühl et al., 1994). Another study suggested that fiber optics only measure changes in the first layer of soil particles that are in direct contact with the sensors (Apitz et al., 1992). Assuming a cylindrical measurement volume, the effective light penetration into the soil would be 
27 mm3 (thick FOMP) and 15 mm3 (thin FOMP) for the two probe sizes described above. This penetration volume will vary with light intensity and soil properties such as texture, water content, and organic matter content. However, considering the work of Apitz et al. (1992), the true measurement volume would be much less than this and would depend on the thickness of the first layer of soil particles adjacent to the probe. As such we report volumetric water contents as cubic millimeters per cubic millimeter to stress the measurement scale of the probe.
Dry and Saturated Soil Reflectance Variability
Ten replicates of dry and saturated soil were measured using a single FOMP. Eight soils of various particle-size distributions were used to examine the influence of soil texture on FOMP measurement (Table 1)
. Each soil was poured into a volume of 60 cm3 and a FOMP was inserted into the soil 10 times to obtain the dry background measurements. A saturation paste was then made for each soil (Rhoades, 1996), and then the probe was again inserted ten times to measure the saturated soil reflectance.
View this table:
[in this window]
[in a new window]
|
Table 1 Coefficients of variation (CV) of air-dry and saturated background light intensity readings for different soils arranged in decreasing sand content
|
|
Fiber Optic Mini-Probe Soil Moisture Calibration Chamber
To examine the quantitative relationship between the output light signal and the soil water content, we designed a special soil column where we could pack air-dried soil and then increase its water content as uniformly as possible in small incremental steps (Fig. 1b). The cylindrical chamber, which is made of dark polyvinyl chloride material, is light proof, and holds 13.6 cm3 of soil. The chamber has four inlets for the insertion of hypodermic needles for water input, and four inlets for the insertion of fiber optic mini-probes. A number of small holes in the lid were made to serve both as air vents and to allow excess water to seep out of the chamber once the soil saturates. The injection needles were inserted 0.44 cm radially from the edge to the center of mass of each quarter. The probes were inserted 1 cm radially from the edge at two depths: 1 cm above and below the injection needles (Fig. 1b).
Creating a system with a homogeneous water content at a 10- to 20-mm3 scale is quite a challenge. Some of the problems include a vertical gradient in water content as well as spatial heterogeneities due to differences in particle arrangement and bulk density. The calibration chamber shown in Fig. 1b was designed to minimize these issues. The spatial orientation of the probes enabled measurements at two depths within the chamber. Furthermore, two probe sizes were used and water was applied to the center of each quadrant within the chamber.
Fiber Optic Mini-Probe Calibration
Three soils were used for our preliminary experiments to construct and test calibrations of the FOMPs to measure soil water content. These soils were silica sand (90.5% SiO2), Delhi sand (mixed, thermic, Typic Xeropsamment), and Botella clay loam (fine-loamy, mixed, superactive, thermic Pachic Argixeroll) (Table 2)
.
After a soil was packed into the calibration chamber (Fig. 1b) to a known bulk density, we calculated the total pore space and thereby the total volume of water required to saturate the soil. The total saturation volume was then divided into a number of equal increments (usually five) and these increments became the steps for each water application event in the experiment. Using a precise multichannel peristaltic pump (Ismatec-IPC, Ismatec, Zurich, Switzerland)1
we delivered each water increment at a rate of 26 mL h-1 (corresponding with a 3% change in volumetric water content in 1 min). The injection channels were individually calibrated to account for any differences in flow. We allowed water redistribution to occur and waited until the light intensity signal stabilized before adding additional water. Experiments were performed six times for the silica sand and four times each for the Delhi sand and Botella clay loam. The calibration curve developed for each soil type represents an average of all data points at each established water content for a specific probe.
Calibration Verification Experiments
To test the calibration method we packed soils into a small column (3 cm3) (Fig. 1c). Initially the sampling window was closed and sealed with black tape and the small FOMP was inserted from the bottom. Probe 4 (a thin probe calibrated in the upper position of the calibration chamber, Fig. 1b) and its resultant calibration curve was used. This probe was chosen since it had the largest coefficient of variation (CV) of the four probes. The column was vertically aligned and homogeneously packed to the same bulk densities used in the calibration chamber. The dry soil background was then measured. Next, a known volume of water was added to the column and water redistribution occurred. After the stabilized light intensity was recorded, a soil sample was removed from the sampling window. This was done by opening the window and removing an 3-mm-thick disk-shaped sample directly adjacent to the FOMP tip using a rectangular tipped spatula. This sample was used to measure gravimetric water content. Volumetric water content (
V) was determined both from the known volume of water added to the system and converted from the gravimetric water content measurement. The V determined from the gravimetric measurements was used since this method produced more reproducible data. The experiment was replicated 35 times for both the silica sand and Delhi sand and 29 times for the Botella clay loam.
|
Results and discussion
|
|---|
Soil Reflectivity–Water Content Relationship
Fiber optic mini-probes were used in the calibration chamber to monitor changes in light intensity as increasing water content decreased the soil reflectivity (Fig. 2)
. In Fig. 2, the values printed under the steps are the exact cumulative volumetric water contents for each step established with the multichannel peristaltic pump. These stated volumetric water contents are the volume-averaged water contents of the entire chamber and not necessarily the water content immediately in front of the probe. Since output light intensity was monitored in real time, we were able to allow adequate redistribution time following each wetting front for all the steps, as evidenced by the stable intensity signals at each step in Fig. 2.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2 Normalized intensity response to water additions in the calibration chamber of two probes (thin upper and lower) for (a) silica sand, (b) Delhi sand, and (c) Botella clay loam
|
|
The wetting fronts in Fig. 2 also illustrate other information regarding water redistribution. In the silica sand, the injections at low water contents display a valley at the beginning of the injection (Fig. 2a and 3b)
. These valleys are probably due to the water injection rate exceeding its redistribution rate to the surrounding volume in the chamber. Then, as redistribution occurs, the signals stabilize at a higher intensity. This was only seen in the upper probes (Probes 3 and 4) at water contents <0.14 mm3 mm-3 and is reasonable because of the coarse-textured nature of the silica sand. This shows the highly sensitive ability of the FOMPs to measure small changes in water content in such a small volume.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 3 The relationship between normalized light intensity and volumetric water content. (a) The resulting relationship for the experiment shown in Fig. 2a. (b) A separate experiment which illustrates the relationship at volumentric water content (V) <0.7 mm3 mm-3. (c) The resulting relationship for the experiment in Fig. 3b
|
|
The wetting fronts in Fig. 2 were converted to a plot of normalized output light intensity vs. volumetric water content to develop a moisture calibration curve (Fig. 3a). A nonlinear relationship exists between the normalized intensity and water content (Fig. 3a). The first increment of water addition caused a relatively large change in the normalized output light intensity, decreasing it from unity to 0.72, whereas the subsequent increments of water reduced the signal from 0.72 to 0.47. However, by excluding the initial dry data point, a linear relationship can be used with a correlation coefficient (r2) >0.98 (Fig. 3a). One explanation for the nonlinearity and large initial change in water content is that since the first portion of any amount of water added to a soil is adsorbed onto the solid phase surfaces, a relatively large change in the initial soil reflective properties occurs. Afterward, additional water simply increases the thickness of this established water film on the soil surfaces until all the pores fill with water and the soil saturates.
To further investigate the relationship between output light intensity and lower water contents we performed another incremental water increase experiment in the calibration chamber with the first water increment divided into three smaller steps
. The step response curve and the calibration curve derived from this experiment are shown in Fig. 3b and 3c, which again illustrate that in the silica sand the first portion of the added water causes the largest relative change in the output light intensity. Again excluding the initial dry data point and applying a linear regression yields an r2 >0.998. Even at a very low moisture range the FOMP is capable of responding to small changes in water content.
Calibration of Fiber Optic Mini-Probes to Measure Water Content
The discussion in the above section pertained to the details of two individual relationships between soil reflectivity and water content for a silica sand as measured using a single FOMP. Overall, we produced 24 calibrations similar to the one described by Fig. 3a for the silica sand (six replicates with four probes per replicate) and 16 calibration curves for the Delhi sand and Botella clay loam each (four replicates with four probes per replicate). All of these replications consistently demonstrated the ability of the system to measure soil water content at a single point in soil. From these calibrations, we averaged the replicates for a single probe to create a calibration curve for each soil and that specific probe. We used an individual calibration curve because of probe-to-probe variability. This is evidenced in Table 3
by comparing average intensity values between the probes. We also suggest developing the calibration curves from a series of replicates, particularly in coarser soils, because of differences in particle arrangements between packing.
Table 3 summarizes the data for the 56 individual calibration curves that were measured in this study. The calibration data shown for each of the probes are the averages of either six or four replicated runs in silica sand and the two soils, respectively. While the calibration curves all had r2 >0.98, the data in Table 3 provides an overall picture of the degree of difficulty that such factors as the soil type, individual probes, and most importantly the small-scale variability of soil matrix in front of each probe can impose on the development of a unique calibration curve. In the case of the fiber optic mini-probe described here, one important source of variability may be the differences between the actual volumetric water content of the small soil volume in front of the probe and the assumed water content which is the volume-averaged water content of the chamber. Although Probes 1 and 3 had thicker fiber optic bundle diameters than Probes 2 and 4, and thus a slightly larger sampling volume, there appears to be little difference in the CVs related to the size of the FOMP (Table 3).
Alternative Calibration Strategy
Another method that utilizes the FOMP's in situ and small-scale capabilities is by calibrating the output light intensity against the soil saturated reflectivity. In this way, changes in intensity can be related to the saturated soil reflectivity to produce a measurement of the percentage of saturation. The linear relationship between soil reflectivity and water content to low levels of V (at least 0.07 mm3 mm-3) makes this method extremely attractive when saturated conditions can be produced around the soil probe. In addition the highly sensitive response of the FOMP can be quite useful when knowledge of small changes in water content is desired.
Effect of Soil Texture on Reflectivity–Moisture Content Relationship
Figure 4
shows the moisture calibration curves for each soil generated by averaging across all the thin probes and all replicates. The two sandy soils, despite their very different moisture calibration curves, have similar moisture response patterns, namely, a large change in the output signal with the first increment of water and then a uniform and somewhat linear response. The moisture response pattern of the Botella clay loam, which is characterized by a much smaller initial moisture response and some specific patterns thereafter, appears quite different in comparison with the two sands. It takes approximately three times more water in the Botella clay loam than in the sands to cause an equivalent signal change. Compared with the two sandy soils, the much higher surface area of the Botella clay loam (among other factors) might be the reason for the smaller magnitude of the first portion of the calibration curve between the air-dry state and the lowest water content step (Fig. 2 and 4).
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 4 The impact of soil type on the relationship between volumetric water content and light reflectance for the thin probes (upper and lower, Probes 2 and 4). Error bars represent standard deviation
|
|
Variability in Dry Soil Reflectivity
In a dry soil, the output light intensity is simply a function of the bulk reflective properties of the soil matrix in front of the probe. Differences in reflectivity between soils are due to soil color, texture, abundance and spatial distribution of organic C, and particle-size distribution. Despite the differences in the reflective properties (due to texture and color) of the silica sand and the other seven soils, the CV of the output light intensity for the soils averaged 1.8% (1.1–2.4%, Table 1). These small CVs suggest that differences in these soils and subsequent orientation of organic and mineral particles around the probe do not produce a great deal of error in soil reflectivity.
Variability in Saturated Soil Reflectivity
The CV for the saturated soils ranged from 0.4% for a silt loam to 5% for the silica sand with an average of 2.4% (Table 1). Examination of the textures of the soils in Table 1 indicates that saturated reflectivity CV decreases as the texture becomes finer. One explanation for this is the greater abundance of micropores in fine-textured soils resulting from their wider particle-size distributions. The small fiber optic probe used in these measurements may encounter less spatial variability in the homogeneous fine textured soils relative to the more coarse soils. Therefore, a probe larger than those in this study may not encounter this variability due to an adequate representation of the particle-size distribution of the soil across the probe's surface area.
Another factor that contributes to the dry and saturated soil reflectivity CV is the probe insertion. Certainly, the small CV values of the dry soil reflectivity suggest that probe insertion does not affect dry measurements. However when the soil is wet, variation in contact between the surrounding soil and the FOMP may contribute to the measured CV. Soil texture may again influence this result since finer-textured soils with greater surface area are expected to produce a greater contact with the probe. Nonetheless, none of the measured CVs for any of the three soils exceeded 5%.
Verification of Calibration Curves
For any new method it is important to stringently test the system. Using the FOMP and the small soil column (Fig. 1c) we performed 35 (silica sand), 35 (Delhi sand), and 29 (Botella clay loam) individual measurements of moisture content where each measurement represents a separate packing, dry soil reflectivity measurement, water addition, and wet soil reflectivity measurement. Comparison of these measurements with the previously obtained calibration curves shows that the independent measurements generally fit within the calibration curve's 95% confidence intervals (Fig. 5) . Goodness of fit between the measurements and the calibration curve was 0.92 for both the Botella clay loam and Delhi sand and 0.64 for the silica sand.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5 Independent measurements of volumetric water content compared with Probe 4 calibration curves and 95% confidence intervals for (a) silica sand, (b) Delhi sand, and (c) Botella clay loam
|
|
The results indicate that the FOMPs used in these experiments have greater accuracy in the Botella clay loam and Delhi sand than in the silica sand. We propose that this difference is largely a result of the lack of fine-textured material in the silica sand relative to the Delhi sand and Botella clay loam. The coarse nature of the silica sand has two effects on the measurement of moisture content. The first is that contact between the coarse grains and the fiber optic probes is more variable. This produces a greater water content heterogeneity at the probe's measurement scale. The second effect is that the coarse grains have a more heterogeneous particle arrangement at the scale of the probes used in this study. In light of this, FOMPs with larger surface areas may reduce the measurement errors by capturing a larger picture of the particle-size distribution.
|
Summary and conclusions
|
|---|
In this study we examined the feasibility of using FOMPs of 3.4- and 2.5-mm diameter for real time in situ measurement of water content within small soil volumes. Similar to other existing soil moisture sensors, the use of this technique requires developing a corresponding calibration curve that describes the measured relationship. For this system the measurement consists of a relationship between volumetric water content and the intensity of light reflected back from the soil.
When considering whether to use this device, a preliminary test should be performed to ensure that the FOMP size yields low variability with respect to the soil's dry and saturated reflectivity. Our results indicated that the FOMP sizes used in this study performed better with finer-textured soils. Also, the stability in the dry and saturated soil light reflectivity measurements (CV < 5%) suggested that a consistent and statistically significant correlation can be developed.
We have suggested two calibration methods for the FOMP. The first method targeted absolute water content. Replicated calibrations should be performed for each individual FOMP and a specific soil. Using this methodology our linear calibration curves had r2 >0.98 for all three soils studied. The second method used the linear nature of this calibration to suggest the ability of the FOMP to measure water contents relative to the saturated water content. Output light intensity measured by the FOMP can be normalized against the saturated soil light reflectivity to obtain a water content relative to the saturated water content.
The FOMP system's major advantages are that it measures water content at very small scales (15–20 mm3) with high temporal resolution and is very sensitive to small changes in water content. In addition, the system can be multiplexed to give a spatial water content distribution. Insertion of probes is simple and minimally disturbs the soil.
At this time the use of the FOMP as a soil moisture measurement device has two limitations. The first is that the calibration requires an extremely accurate knowledge of the water content directly in front of the probe. Even in small soil volumes it is difficult to establish a known uniform water content at the measurement scale of the FOMP. The second limitation is the FOMP's apparent decrease in performance in coarse-textured porous media. Additional studies using probes with larger surface areas may improve these factors.
The FOMP system may be ideal in situations where water content is dynamic and changing at small spatial scales, especially where these changes have a large impact on other processes. Some examples of possible applications are studying the effect of changing water content on film flow at the macropore–matrix interface, sorption processes, microbial activity (which is highly depended on water content), and rhizosphere dynamics. Even without careful calibration, the FOMP system can be used to monitor the arrival and passage of wetting fronts in any soil system. With careful calibration, the FOMP system can be a valuable device for measuring absolute water content as well as for extremely accurate measurements of relative water content.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank Chris Campbell for his review and insights into this manuscript. We would also like to thank Johnny Liu for his laboratory assistance. The first author thanks the Ministerio de Educación y Cultura (Spain) for financial support.
|
NOTES
|
|---|
1 This product is available through different suppliers. No endorsement is intended. 
Received for publication April 13, 1998.
|
REFERENCES
|
|---|
- Alessi R.S., Prunty L. Soil water determination using fiber optics. Soil Sci. Soc. Am. J. 1986;50:860-863.[Abstract/Free Full Text]
- Amato M., Ritchie J.T. Small spatial scale soil water content measurement with time-domain reflectometry. Soil Sci. Soc. Am. J. 1995;59:325-329.[Abstract/Free Full Text]
- Apitz S.E., Borbridge L.M., Theriault G.A., Lieberman S.H. Remote in situ determination of fuel products in soils: Field results and laboratory investigations. Analusis 1992;20:461-474.
- Consentino P., Grossman B., Shieh C., Doi S., Xi H., Erbland P. Fiber optic chloride sensor development. J. Geotech. Eng. 1995;121:610-617.
- Ghodrati M. Point measurement of solute transport processes in soil using fiber optic sensors. Soil Sci. Soc. Am. J. 1999;63:471-479.[Abstract/Free Full Text]
- Goyet, C., D.R. Walt, and P.G. Brewer. 1992. Development of fiber optic sensor for measurement of pCO2 in sea water: Design criteria and sea trials. Deep-sea Research Part A-Oceanographic Res. Papers. 39(6A):1015–1026.
- Hadjiloucas S., Karatzas L.S., Keating D.A., Usher M.J. Optical sensors for monitoring water uptake in plants. J. Lightwave Technol. 1995;13:1421-1428.
- Krohn, D.A. 1988. Chemical analysis. p. 193–209. Fiber optic sensors: Fundamentals and applications. Instrument Society of America, Research Triangle Park, NC.
- Kühl M., Lassen C., Jørgensen B.B. Light penetration and light intensity in sandy marine sediments measured with irradiance and scalar irradiance fiber-optic miroprobes. Mar. Ecol. Prog. Ser. 1994;105:139-148.
- Maher M.H., Shahriari M.R. A fiber optic chemical sensor for measurement of groundwater pH. J. Test. Eval. 1993;21:448-452.
- Muto, S., A. Fukasawa, T. Ogawa, M. Morisawa, and H. Ito. 1990. Optical detection of moisture in air and in soil using dye-doped plastic fibers. Jpn. J. Appl. Physics Part 2-letters. 29(6):L1023–L1025.
- Nielsen D.R., Kutilek M., Parlange M.B. Surface soil water content regimes—Opportunities in soil science. J. Hydrol. 1996;182:32-55.
- Nissen H.H., Moldrup P., Henriksen K. High-resolution time domain reflectometry coil probe for measuring soil water content. Soil Sci. Soc. Am. J. 1998;62:1203-1211.[Abstract/Free Full Text]
- Reece C.F. Evaluation of a line heat dissipation sensor for measuring soil matric potential. Soil Sci. Soc. Am. J. 1996;60:1022-1028.[Abstract/Free Full Text]
- Rhoades J.D. Salinity: Electrical conductivity and total dissolved solids. In: Sparks D.L., ed. Methods of soil analysis. Madison, WI: Part 3. SSSA Book Ser. 5. SSSA, 1996:417-435.
- Rogers K.R., Poziomek E.J. Fiber optic sensors for environmental monitoring. Chemosphere 1996;33:1151-1174.
- Rudolph D.L., Kachanoski G., Celia M.A., LeBlanc D.R., Stevens J.H. Infiltration and solute transport experiments in unsaturated sand and gravel, Cape Cod, Massachusetts: Experimental design and overview of results. Water Resour. Res. 1996;33:519-532.
- Tokunaga T., Wan J. Water film flow along fracture surfaces of porous rock. Water Resour. Res. 1997;33:1287-1295.
- Topp G.C., Watt M., Hayoe H.N. Point specific measurement and monitoring of soil water content with an emphasis on TDR. Can. J. Soil Sci. 1996;76:307-316.
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]
|
|
|
|
|
|
|
F. Garrido, M. Ghodrati, C. G. Campbell, and M. Chendorain
Detailed Characterization of Solute Transport in a Heterogeneous Field Soil
J. Environ. Qual.,
March 1, 2001;
30(2):
573 - 583.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
