Adv. Radio Sci., 6, 1–4, 2008www.adv-radio-sci.net/6/1/2008/© Author(s) 2008. This work is distributed underthe Creative Commons Attribution 3.0 License.

Advances inRadio Science

Time domain reflectrometry measurements using a movableobstacle for the determination of dielectric profiles

B. Will, M. Gerding, S. Schultz, and B. Schiek

Ruhr-University Bochum, Institute of High Frequency Engineering, Universitatsstr. 150, 44801 Bochum, Germany

Abstract. Microwave techniques for the measurement of thepermittivity of soils including the water content of soils andother materials, especially TDR (time domain reflectometry),have become accepted as routine measurement techniques.This summary deals with an advanced use of the TDR prin-ciple for the determination of the water content of soil along aprobe. The basis of the advanced TDR technique is a waveg-uide, which is inserted into the soil for obtaining measure-ments of the effective soil permittivity, from which the watercontent is estimated, and an obstacle, which can mechani-cally be moved along the probe and which acts as a refer-ence reflection for the TDR system with an exactly knownposition. Based on the known mechanical position of thereference reflection, the measured electrical position can beused as a measure for the effective dielectric constant of theenvironment. Thus, it is possible to determine the effectivedielectric constant with a spatial resolution given by the stepsize of the obstacle displacement.

A conventional industrial TDR-system, operating in thebaseband, is used for the signal generation and for the evalu-ation of the pulse delay time of the obstacle reflection. Thus,a cost effective method for the acquisition of the dielectricmeasurement data is available.

1 Introduction

The aim of this investigation is to obtain representative pro-file measurement results from a material under test. Time do-main reflectometry (TDR) is known as a method to obtain areliable estimation of the soil water content from a measure-ment of the real partε′ as well as the imaginary partε′′ ofthe relative permittivityεr (Kupfer, 2005; Hoekstra, 1975).In many cases TDR is used in order to get just one integralvalue for the relative permittivity of the material (Robinson,2000). In this case the reflection from the end of the probe is

Correspondence to:B. Will(bianca.will@rub.de)

2 B. Will et al.: Dielectric Profile Measurements

e1

e2

e3

e4

e5

e6

lmech

lmech

lmech

lmech

Fig. 1. Functional principle

The measured pulse delay time is proportional to the relativeeffective permittivity εeff of the penetrated medium aroundthe analyzed probe length lmech:

∆tmeas =2lmech

c0

√εeff . (2)

By comparison with the theoretical pulse delay time ∆tref :

∆tref =2lmech

c0(3)

the effective permittivity along the analyzed probe lengthlmech can be determined:

εeff =(

∆tmeas

∆tref

)2

. (4)

Obviously the spatial resolution of the system directly de-pends on the length interval lmech and on the pulse delaytime ∆tref corresponding to the obstacle displacement, re-spectively. The higher the spatial resolution the shorter thepulse delay time. If the pulse delay time is in the order of theinherent time jitter ∆tjitter of the TDR system, the upperresolution limit is reached and can be determined by:

∆lmin =12c0∆tjitter . (5)

The minimum resolution length is a multiple of ∆lmin foraccurate measurement results. To keep the estimated errorbelow 1%, the practical minimum resolution length will beapproximately 100∆lmin. In relation to this, the effectivetime jitter of±500 fs without an averaging of the TDR proto-type system data would lead to a minimum resolution lengthof 30mm.

The basic functional principle of the measuring system isillustrated in Fig. 1.

In this case the movable obstacle is realized as a movableshort circuit. This short is moved stepwise by lmech and ineach position the pulse delay time is measured. With thispulse delay time and the pulse delay time measured withouta dielectric surrounding one gets the effective dielectric con-stant for each step. Thus, one can perform a measurementfor the characterization of dielectric profiles with a spatialresolution given by the step size with which the obstacle ismoved.

movable

short circuit

cable duct

two-wire line

Fig. 2. Probe design and field distribution

3 Probe Design

In order to obtain good measurement results it is very im-portant to design a probe, which allows measurements in anenvironment with high permittivity values on the one handand high resolution measurements on the other hand (Knight,1992; Zegelin, 1989). Thus it is necessary to have a goodinteraction between the electromagnetic field and the envi-ronment. On the other hand it is necessary to have a certaingap between the conductors and the environment, otherwisehigh dielectric constants in the environment would cause astrong reflection. Due to these requirements the probe shownin Fig. 2 was designed.

This two-wire probe with a short as a movable obstacleis a good compromise between maximum interaction of theelectromagnetic field and the environment on the one handand the detection of high permittivities on the other hand.

4 Measurement Results

For first measurements the obstacle was placed consecutivelyin two positions as shown in Fig. 3. In a first step the delaytime in position 1 was measured as a reference. Then the ob-stacle was moved into the second position and the delay timewith air as environment was measured. Then the cavity be-tween position 1 and position 2 was consecutively filled withdifferent materials. The delay time was measured for eachmaterial and compared with the delay time without filling.For these measurements the obstacle was fixed in position2. Thus measurement errors caused by a displacement of theobstacle could not occur.

Fig. 4 shows the measurement results for a fixed obstacleposition. The cavity between position 1 and position 2 wasfilled with five different materials. For each material the de-

Fig. 1. Functional principle.

measured and converted into a value for the relative permit-tivity. This summary deals with a method that uses differentreflections. Thus, it is possible to measure a dielectric pro-file along the probe. The different reflections are caused byan obstacle which is moved mechanically along the probe.The step size of the motion of this obstacle is one importantparameter for the possible spatial resolution of the system.

2 Functional principle

The propagation speed of electromagnetic waves is directlyrelated to the effective permittivity of the penetrated medium.In closed waveguide structures like in circular waveguides orcoaxial waveguides with a relative permittivityεr the propa-gation velocity is given by:

c =c0

√εr

, (1)

while the propagation velocity along the probe of the testsetup shown in Fig.1 is determined by the effective permit-tivity ε0 ·εeff. The effective permittivity describes the permit-tivity of the dielectric mixture of the penetrated medium. Thevolume of e.g. soil for which the estimation is valid, primar-ily depends on the design of the probe (Heimovaara, 1993).By moving a reflecting obstacle stepwise along the probe, apermittivity profile can be determined by using the relationbetween the free space pulse delay time1tref correspond-ing to the mechanical lengthlmech of the obstacle displace-

Published by Copernicus Publications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V.

2 B. Will et al.: Dielectric profile measurements2 B. Will et al.: Dielectric Profile Measurements

e1

e2

e3

e4

e5

e6

lmech

lmech

lmech

lmech

Fig. 1. Functional principle

The measured pulse delay time is proportional to the relativeeffective permittivity εeff of the penetrated medium aroundthe analyzed probe length lmech:

∆tmeas =2lmech

c0

√εeff . (2)

By comparison with the theoretical pulse delay time ∆tref :

∆tref =2lmech

c0(3)

the effective permittivity along the analyzed probe lengthlmech can be determined:

εeff =(

∆tmeas

∆tref

)2

. (4)

Obviously the spatial resolution of the system directly de-pends on the length interval lmech and on the pulse delaytime ∆tref corresponding to the obstacle displacement, re-spectively. The higher the spatial resolution the shorter thepulse delay time. If the pulse delay time is in the order of theinherent time jitter ∆tjitter of the TDR system, the upperresolution limit is reached and can be determined by:

∆lmin =12c0∆tjitter . (5)

The minimum resolution length is a multiple of ∆lmin foraccurate measurement results. To keep the estimated errorbelow 1%, the practical minimum resolution length will beapproximately 100∆lmin. In relation to this, the effectivetime jitter of±500 fs without an averaging of the TDR proto-type system data would lead to a minimum resolution lengthof 30mm.

The basic functional principle of the measuring system isillustrated in Fig. 1.

In this case the movable obstacle is realized as a movableshort circuit. This short is moved stepwise by lmech and ineach position the pulse delay time is measured. With thispulse delay time and the pulse delay time measured withouta dielectric surrounding one gets the effective dielectric con-stant for each step. Thus, one can perform a measurementfor the characterization of dielectric profiles with a spatialresolution given by the step size with which the obstacle ismoved.

movable

short circuit

cable duct

two-wire line

Fig. 2. Probe design and field distribution

3 Probe Design

In order to obtain good measurement results it is very im-portant to design a probe, which allows measurements in anenvironment with high permittivity values on the one handand high resolution measurements on the other hand (Knight,1992; Zegelin, 1989). Thus it is necessary to have a goodinteraction between the electromagnetic field and the envi-ronment. On the other hand it is necessary to have a certaingap between the conductors and the environment, otherwisehigh dielectric constants in the environment would cause astrong reflection. Due to these requirements the probe shownin Fig. 2 was designed.

This two-wire probe with a short as a movable obstacleis a good compromise between maximum interaction of theelectromagnetic field and the environment on the one handand the detection of high permittivities on the other hand.

4 Measurement Results

For first measurements the obstacle was placed consecutivelyin two positions as shown in Fig. 3. In a first step the delaytime in position 1 was measured as a reference. Then the ob-stacle was moved into the second position and the delay timewith air as environment was measured. Then the cavity be-tween position 1 and position 2 was consecutively filled withdifferent materials. The delay time was measured for eachmaterial and compared with the delay time without filling.For these measurements the obstacle was fixed in position2. Thus measurement errors caused by a displacement of theobstacle could not occur.

Fig. 4 shows the measurement results for a fixed obstacleposition. The cavity between position 1 and position 2 wasfilled with five different materials. For each material the de-

Fig. 2. Probe design and field distribution.

ment and the measured delay time1tmeascorresponding tothe electrical length and the evaluation of the TDR signal.The measured pulse delay time is proportional to the relativeeffective permittivityεeff of the penetrated medium aroundthe analyzed probe lengthlmech:

1tmeas=2lmech

c0

√εeff . (2)

By comparison with the theoretical pulse delay time1tref:

1tref =2lmech

c0(3)

the effective permittivity along the analyzed probe lengthlmechcan be determined:

εeff =

(1tmeas

1tref

)2

. (4)

Obviously the spatial resolution of the system directly de-pends on the length intervallmech and on the pulse delaytime 1tref corresponding to the obstacle displacement, re-spectively. The higher the spatial resolution the shorter thepulse delay time. If the pulse delay time is in the order ofthe inherent time jitter1tjitter of the TDR system, the upperresolution limit is reached and can be determined by:

1lmin =1

2c01tjitter . (5)

The minimum resolution length is a multiple of1lmin for ac-curate measurement results. To keep the estimated error be-low 1%, the practical minimum resolution length will be ap-proximately 1001lmin. In relation to this, the effective timejitter of ±500 fs without an averaging of the TDR prototypesystem data would lead to a minimum resolution length of30 mm.

The basic functional principle of the measuring system isillustrated in Fig.1.

B. Will et al.: Dielectric Profile Measurements 3

Fig. 3. Measurement setup for a fixed obstacle position

1.6

1.8

2

2.2

2.4

2.6

2.8

3

effe

ctive

pe

rmittivity

sand 1

dry

empty empty

gravelgravel

sand 2

dry

sand 1

moist

1 2 3 4 5 6 7 8

sand 1

dry

cavity fillings

Fig. 4. Measurement results for a fixed obstacle position andindividual cavity fillings

lay time was measured five times. The effective permittivityof each material was calculated by using equation 4.

Thus, these first measurement results show that the com-parison of delay time intervalls is a very useful method forthe characterization of effective permittivities.In a next step a measurement with a movable obstacle, in thiscase a short, was performed. Fig. 5 a) shows the measure-ment setup.

A tube with a diameter of 20 cm was divided into fourcavities, which are filled with different types of gravel andsand. Fig. 5 b) shows the measurement results as a color bar.The two-wire probe is located in the radial center of the tube.The dielectric profile of the environment was measured witha step size of 2.5 cm and the length of each cavity was 25 cm.Fig. 6 shows the measurement results.

We observe some deviations concerning the effective per-mittivity, which can partially be explained by the manualmoving of the obstacle. Nevertheless the different materi-

Fig. 5. Measurement setup with a movable obstacle (a) and mea-surement results as a color bar (b)

0 0.5 1 1.51

1.5

2

2.5

3

3.5

obstacle position [m]

effective p

erm

ittivity

2

Fig. 6. Measurement results with a movable obstacle

als are distinguishable from each other.In order to reduce these uncertainties, the obstacle has to beoptimized. For certain measurement results it is very im-portant to know the position of the obstacle exactly. If theobstacle is realized as a short circuit, one should ensure, thatthe obstacle does not tilt between the two transmission lines.Thus, the obstacle which is moved should be very thin andit should provide a good contact between the transmissionlines.Furthermore the obstacle should be moved automatically toreduce uncertainties during the positioning.

Fig. 7 shows first results for an automated measurementwith a proper short circuit. In this case every other cavitywas filled with dry sand. This figure shows that there areonly small uncertainties and the different cavities are welldefined. The transmission lines are fixed by three pairs ofteflon pads, which cause small deviations.In a next step the separators between the different cavitiesare removed. The entire tube was filled with dry sand andin several positions water was locally added. Fig. 8 showsthe measurement results with one and with two moist areas.

Fig. 3. Measurement setup for a fixed obstacle position.

In this case the movable obstacle is realized as a movableshort circuit. This short is moved stepwise bylmech and ineach position the pulse delay time is measured. With thispulse delay time and the pulse delay time measured withouta dielectric surrounding one gets the effective dielectric con-stant for each step. Thus, one can perform a measurementfor the characterization of dielectric profiles with a spatialresolution given by the step size with which the obstacle ismoved.

3 Probe Design

In order to obtain good measurement results it is very im-portant to design a probe, which allows measurements in anenvironment with high permittivity values on the one handand high resolution measurements on the other hand (Knight,1992; Zegelin, 1989). Thus it is necessary to have a goodinteraction between the electromagnetic field and the envi-ronment. On the other hand it is necessary to have a certaingap between the conductors and the environment, otherwisehigh dielectric constants in the environment would cause astrong reflection. Due to these requirements the probe shownin Fig. 2 was designed.

This two-wire probe with a short as a movable obstacleis a good compromise between maximum interaction of theelectromagnetic field and the environment on the one handand the detection of high permittivities on the other hand.

4 Measurement results

For first measurements the obstacle was placed consecutivelyin two positions as shown in Fig.3. In a first step the delaytime in position 1 was measured as a reference. Then the ob-stacle was moved into the second position and the delay time

Adv. Radio Sci., 6, 1–4, 2008 www.adv-radio-sci.net/6/1/2008/

B. Will et al.: Dielectric profile measurements 3

B. Will et al.: Dielectric Profile Measurements 3

Fig. 3. Measurement setup for a fixed obstacle position

1.6

1.8

2

2.2

2.4

2.6

2.8

3

effective p

erm

ittivity

sand 1

dry

empty empty

gravelgravel

sand 2

dry

sand 1

moist

1 2 3 4 5 6 7 8

sand 1

dry

cavity fillings

Fig. 4. Measurement results for a fixed obstacle position andindividual cavity fillings

lay time was measured five times. The effective permittivityof each material was calculated by using equation 4.

Thus, these first measurement results show that the com-parison of delay time intervalls is a very useful method forthe characterization of effective permittivities.In a next step a measurement with a movable obstacle, in thiscase a short, was performed. Fig. 5 a) shows the measure-ment setup.

A tube with a diameter of 20 cm was divided into fourcavities, which are filled with different types of gravel andsand. Fig. 5 b) shows the measurement results as a color bar.The two-wire probe is located in the radial center of the tube.The dielectric profile of the environment was measured witha step size of 2.5 cm and the length of each cavity was 25 cm.Fig. 6 shows the measurement results.

We observe some deviations concerning the effective per-mittivity, which can partially be explained by the manualmoving of the obstacle. Nevertheless the different materi-

Fig. 5. Measurement setup with a movable obstacle (a) and mea-surement results as a color bar (b)

0 0.5 1 1.51

1.5

2

2.5

3

3.5

obstacle position [m]

effective p

erm

ittivity

2

Fig. 6. Measurement results with a movable obstacle

als are distinguishable from each other.In order to reduce these uncertainties, the obstacle has to beoptimized. For certain measurement results it is very im-portant to know the position of the obstacle exactly. If theobstacle is realized as a short circuit, one should ensure, thatthe obstacle does not tilt between the two transmission lines.Thus, the obstacle which is moved should be very thin andit should provide a good contact between the transmissionlines.Furthermore the obstacle should be moved automatically toreduce uncertainties during the positioning.

Fig. 7 shows first results for an automated measurementwith a proper short circuit. In this case every other cavitywas filled with dry sand. This figure shows that there areonly small uncertainties and the different cavities are welldefined. The transmission lines are fixed by three pairs ofteflon pads, which cause small deviations.In a next step the separators between the different cavitiesare removed. The entire tube was filled with dry sand andin several positions water was locally added. Fig. 8 showsthe measurement results with one and with two moist areas.

Fig. 4. Measurement results for a fixed obstacle position and indi-vidual cavity fillings.B. Will et al.: Dielectric Profile Measurements 3

Fig. 3. Measurement setup for a fixed obstacle position

1.6

1.8

2

2.2

2.4

2.6

2.8

3

effe

ctive

pe

rmittivity

sand 1

dry

empty empty

gravelgravel

sand 2

dry

sand 1

moist

1 2 3 4 5 6 7 8

sand 1

dry

cavity fillings

Fig. 4. Measurement results for a fixed obstacle position andindividual cavity fillings

lay time was measured five times. The effective permittivityof each material was calculated by using equation 4.

Thus, these first measurement results show that the com-parison of delay time intervalls is a very useful method forthe characterization of effective permittivities.In a next step a measurement with a movable obstacle, in thiscase a short, was performed. Fig. 5 a) shows the measure-ment setup.

A tube with a diameter of 20 cm was divided into fourcavities, which are filled with different types of gravel andsand. Fig. 5 b) shows the measurement results as a color bar.The two-wire probe is located in the radial center of the tube.The dielectric profile of the environment was measured witha step size of 2.5 cm and the length of each cavity was 25 cm.Fig. 6 shows the measurement results.

We observe some deviations concerning the effective per-mittivity, which can partially be explained by the manualmoving of the obstacle. Nevertheless the different materi-

Fig. 5. Measurement setup with a movable obstacle (a) and mea-surement results as a color bar (b)

0 0.5 1 1.51

1.5

2

2.5

3

3.5

obstacle position [m]

effective p

erm

ittivity

2

Fig. 6. Measurement results with a movable obstacle

als are distinguishable from each other.In order to reduce these uncertainties, the obstacle has to beoptimized. For certain measurement results it is very im-portant to know the position of the obstacle exactly. If theobstacle is realized as a short circuit, one should ensure, thatthe obstacle does not tilt between the two transmission lines.Thus, the obstacle which is moved should be very thin andit should provide a good contact between the transmissionlines.Furthermore the obstacle should be moved automatically toreduce uncertainties during the positioning.

Fig. 7 shows first results for an automated measurementwith a proper short circuit. In this case every other cavitywas filled with dry sand. This figure shows that there areonly small uncertainties and the different cavities are welldefined. The transmission lines are fixed by three pairs ofteflon pads, which cause small deviations.In a next step the separators between the different cavitiesare removed. The entire tube was filled with dry sand andin several positions water was locally added. Fig. 8 showsthe measurement results with one and with two moist areas.

Fig. 5. Measurement setup with a movable obstacle(a) and mea-surement results as a color bar(b).

with air as environment was measured. Then the cavity be-tween position 1 and position 2 was consecutively filled withdifferent materials. The delay time was measured for eachmaterial and compared with the delay time without filling.For these measurements the obstacle was fixed in position2. Thus measurement errors caused by a displacement of theobstacle could not occur.

Figure4 shows the measurement results for a fixed obsta-cle position. The cavity between position 1 and position 2was filled with five different materials. For each material thedelay time was measured five times. The effective permittiv-ity of each material was calculated by using Eq.4.

Thus, these first measurement results show that the com-parison of delay time intervalls is a very useful method forthe characterization of effective permittivities. In a next stepa measurement with a movable obstacle, in this case a short,was performed. Figure5 a) shows the measurement setup.

A tube with a diameter of 20 cm was divided into four cav-ities, which are filled with different types of gravel and sand.Fig. 5 b) shows the measurement results as a color bar. The

B. Will et al.: Dielectric Profile Measurements 3

Fig. 3. Measurement setup for a fixed obstacle position

1.6

1.8

2

2.2

2.4

2.6

2.8

3

effective

perm

ittivity

sand 1

dry

empty empty

gravelgravel

sand 2

dry

sand 1

moist

1 2 3 4 5 6 7 8

sand 1

dry

cavity fillings

Fig. 4. Measurement results for a fixed obstacle position andindividual cavity fillings

lay time was measured five times. The effective permittivityof each material was calculated by using equation 4.

Thus, these first measurement results show that the com-parison of delay time intervalls is a very useful method forthe characterization of effective permittivities.In a next step a measurement with a movable obstacle, in thiscase a short, was performed. Fig. 5 a) shows the measure-ment setup.

A tube with a diameter of 20 cm was divided into fourcavities, which are filled with different types of gravel andsand. Fig. 5 b) shows the measurement results as a color bar.The two-wire probe is located in the radial center of the tube.The dielectric profile of the environment was measured witha step size of 2.5 cm and the length of each cavity was 25 cm.Fig. 6 shows the measurement results.

We observe some deviations concerning the effective per-mittivity, which can partially be explained by the manualmoving of the obstacle. Nevertheless the different materi-

Fig. 5. Measurement setup with a movable obstacle (a) and mea-surement results as a color bar (b)

0 0.5 1 1.51

1.5

2

2.5

3

3.5

obstacle position [m]

effective p

erm

ittivity

2

Fig. 6. Measurement results with a movable obstacle

als are distinguishable from each other.In order to reduce these uncertainties, the obstacle has to beoptimized. For certain measurement results it is very im-portant to know the position of the obstacle exactly. If theobstacle is realized as a short circuit, one should ensure, thatthe obstacle does not tilt between the two transmission lines.Thus, the obstacle which is moved should be very thin andit should provide a good contact between the transmissionlines.Furthermore the obstacle should be moved automatically toreduce uncertainties during the positioning.

Fig. 7 shows first results for an automated measurementwith a proper short circuit. In this case every other cavitywas filled with dry sand. This figure shows that there areonly small uncertainties and the different cavities are welldefined. The transmission lines are fixed by three pairs ofteflon pads, which cause small deviations.In a next step the separators between the different cavitiesare removed. The entire tube was filled with dry sand andin several positions water was locally added. Fig. 8 showsthe measurement results with one and with two moist areas.

Fig. 6. Measurement results with a movable obstacle.4 B. Will et al.: Dielectric Profile Measurements

1.761.411.060.710.36

0

1

2

3

obstacle position [m]

effective p

erm

ittivity

teflon-pads

Fig. 7. Automated measurement with sand in the different cav-ities

0.4 0.6 0.8 1 1.2 1.4 1.6

0

1

2

3

4

effective p

erm

ittivity

obstacle position [m]

0.4 0.6 0.8 1 1.2 1.4 1.6

0

1

2

3

4

obstacle position [m]

effective p

erm

ittivity

water positions

water position

Fig. 8. Measurement results with different moist positions

Thus, this measurement setup is a very proper method for themeasurement of dielectric profiles in the surrounding of theprobe.

5 Conclusions

The results in this summary allow an estimation of the capa-bility of the presented measurement concept for the determi-nation of dielectric profiles and profiles of the water contentof soils, respectively.In addition to well known TDR moisture sensing systems,which only determine the integral value of the moisture con-tent along the sensor, the presented concept is capable of

measuring moisture profiles by use of a conventional in-dustrial TDR-system or a vector network analyzer, respec-tively, for the signal generation and the signal evaluation.To achieve the additional information for reconstructing adielectric profile from the measured data, a special sensoris necessary, which includes a movable reflecting target, aso-called obstacle, which can manually or automatically bemoved along the probe, in order to act as a reference. In thisway a mapping of the effective permittivities is possible.

References

Heimovaara, T.J.: Design of triple wire domain reflectometryprobes in practice and theory, Soil Sci. Soc. Am., 57, 1410–1417,1993.

Hoekstra, P. and Delanay, A.: Dielectric Properties of soils at UHFand microwave frequencies, Journal of Geophysical Research,79, 1699–1708, 1975.

Knight, J.H.: Sensitivity of time domain reflectometry measure-ments to lateral variations in soil water content, Water Res. Re-search, vol. 28, 9, 2345–2352, 1992.

Kupfer, K.: Electromagnetic Aquametry: Electromagnetic WaveInteraction with Water and Moist Substances, Springer Verlag,Berlin, 2005.

Robinson, D.A. and Friedman, S.P.: Parallel Plates Compared withConventional Rods as TDR Waveguides for Sensing Soil Mois-ture, Subsurface Sensing Technologies and Applications Journal,1, 497–511, 2000.

Zegelin, S.J., White, I. and Jenkins, D.R.: Improved field probes forsoil water content and electrical conductivity measurement usingTime Domain Reflectometry, Water Res. Research, 25, 2367–2376, 1989.

Fig. 7. Automated measurement with sand in the different cavities.

two-wire probe is located in the radial center of the tube. Thedielectric profile of the environment was measured with astep size of 2.5 cm and the length of each cavity was 25 cm.Figure6 shows the measurement results.

We observe some deviations concerning the effective per-mittivity, which can partially be explained by the manualmoving of the obstacle. Nevertheless the different materialsare distinguishable from each other. In order to reduce theseuncertainties, the obstacle has to be optimized. For certainmeasurement results it is very important to know the positionof the obstacle exactly. If the obstacle is realized as a shortcircuit, one should ensure, that the obstacle does not tilt be-tween the two transmission lines. Thus, the obstacle whichis moved should be very thin and it should provide a goodcontact between the transmission lines. Furthermore the ob-stacle should be moved automatically to reduce uncertaintiesduring the positioning.

Figure7 shows first results for an automated measurementwith a proper short circuit. In this case every other cavitywas filled with dry sand. This figure shows that there are onlysmall uncertainties and the different cavities are well defined.The transmission lines are fixed by three pairs of teflon pads,which cause small deviations. In a next step the separators

www.adv-radio-sci.net/6/1/2008/ Adv. Radio Sci., 6, 1–4, 2008

4 B. Will et al.: Dielectric profile measurements

4 B. Will et al.: Dielectric Profile Measurements

1.761.411.060.710.36

0

1

2

3

obstacle position [m]

effe

ctive

pe

rmittivity

teflon-pads

Fig. 7. Automated measurement with sand in the different cav-ities

0.4 0.6 0.8 1 1.2 1.4 1.6

0

1

2

3

4

effe

ctive

pe

rmittivity

obstacle position [m]

0.4 0.6 0.8 1 1.2 1.4 1.6

0

1

2

3

4

obstacle position [m]

effe

ctive

pe

rmittivity

water positions

water position

Fig. 8. Measurement results with different moist positions

Thus, this measurement setup is a very proper method for themeasurement of dielectric profiles in the surrounding of theprobe.

5 Conclusions

The results in this summary allow an estimation of the capa-bility of the presented measurement concept for the determi-nation of dielectric profiles and profiles of the water contentof soils, respectively.In addition to well known TDR moisture sensing systems,which only determine the integral value of the moisture con-tent along the sensor, the presented concept is capable of

measuring moisture profiles by use of a conventional in-dustrial TDR-system or a vector network analyzer, respec-tively, for the signal generation and the signal evaluation.To achieve the additional information for reconstructing adielectric profile from the measured data, a special sensoris necessary, which includes a movable reflecting target, aso-called obstacle, which can manually or automatically bemoved along the probe, in order to act as a reference. In thisway a mapping of the effective permittivities is possible.

References

Heimovaara, T.J.: Design of triple wire domain reflectometryprobes in practice and theory, Soil Sci. Soc. Am., 57, 1410–1417,1993.

Hoekstra, P. and Delanay, A.: Dielectric Properties of soils at UHFand microwave frequencies, Journal of Geophysical Research,79, 1699–1708, 1975.

Knight, J.H.: Sensitivity of time domain reflectometry measure-ments to lateral variations in soil water content, Water Res. Re-search, vol. 28, 9, 2345–2352, 1992.

Kupfer, K.: Electromagnetic Aquametry: Electromagnetic WaveInteraction with Water and Moist Substances, Springer Verlag,Berlin, 2005.

Robinson, D.A. and Friedman, S.P.: Parallel Plates Compared withConventional Rods as TDR Waveguides for Sensing Soil Mois-ture, Subsurface Sensing Technologies and Applications Journal,1, 497–511, 2000.

Zegelin, S.J., White, I. and Jenkins, D.R.: Improved field probes forsoil water content and electrical conductivity measurement usingTime Domain Reflectometry, Water Res. Research, 25, 2367–2376, 1989.

Fig. 8. Measurement results with different moist positions.

between the different cavities are removed. The entire tubewas filled with dry sand and in several positions water waslocally added. Figure8 shows the measurement results withone and with two moist areas. Thus, this measurement setupis a very proper method for the measurement of dielectricprofiles in the surrounding of the probe.

5 Conclusions

The results in this summary allow an estimation of the capa-bility of the presented measurement concept for the determi-nation of dielectric profiles and profiles of the water contentof soils, respectively.

In addition to well known TDR moisture sensing systems,which only determine the integral value of the moisture con-tent along the sensor, the presented concept is capable ofmeasuring moisture profiles by use of a conventional in-dustrial TDR-system or a vector network analyzer, respec-tively, for the signal generation and the signal evaluation.To achieve the additional information for reconstructing adielectric profile from the measured data, a special sensoris necessary, which includes a movable reflecting target, aso-called obstacle, which can manually or automatically bemoved along the probe, in order to act as a reference. In thisway a mapping of the effective permittivities is possible.

References

Heimovaara, T. J.: Design of triple wire domain reflectometryprobes in practice and theory, Soil Sci. Soc. Am., 57, 1410–1417,1993.

Hoekstra, P. and Delanay, A.: Dielectric Properties of soils at UHFand microwave frequencies, J. Geophys. Res., 79, 1699–1708,1975.

Knight, J. H.: Sensitivity of time domain reflectometry measure-ments to lateral variations in soil water content, Water Resour.Res., 28, 9, 2345–2352, 1992.

Kupfer, K.: Electromagnetic Aquametry: Electromagnetic WaveInteraction with Water and Moist Substances, Springer Verlag,Berlin, 2005.

Robinson, D. A. and Friedman, S. P.: Parallel Plates Compared withConventional Rods as TDR Waveguides for Sensing Soil Mois-ture, Subsurface Sensing Technologies and Applications Journal,1, 497–511, 2000.

Zegelin, S. J., White, I., and Jenkins, D. R.: Improved fieldprobes for soil water content and electrical conductivity measure-ment using Time Domain Reflectometry, Water Resour. Res., 25,2367–2376, 1989.

Adv. Radio Sci., 6, 1–4, 2008 www.adv-radio-sci.net/6/1/2008/