In situ measurement of soil moisture: a comparison of …jpwalker/papers/jh04.pdfJeffrey P. Walker*, Garry R. Willgoose1, Jetse D. Kalma Department of Civil, Surveying and Environmental

UNCORRECTED PROOF

In situ measurement of soil moisture: a comparison

of techniques

Jeffrey P. Walker*, Garry R. Willgoose1, Jetse D. Kalma

Department of Civil, Surveying and Environmental Engineering, The University of Newcastle, Callaghan, NSW 2308, Australia

Received 3 December 2002; revised 19 January 2004; accepted 23 January 2004

Abstract

A number of automated techniques for point measurement of soil moisture content have been developed to an operational

level over the past few decades. While each of those techniques have been individually evaluated by the thermogravimetric

(oven drying and weighing) method, typically under laboratory conditions, there have been few studies which have made a

direct comparison between the various techniques, particularly under field conditions. This paper makes an inter-comparison of

the Virribw, Campbell Scientific CS615 reflectometer, Soil Moisture Equipment Corporation TRASEw buriable- and

connector-type time domain reflectometry (TDR) soil moisture sensors, and a comparison of the connector-type TDR sensor

with thermogravimetric measurements for data collected during a 2-year field study. Both qualitative and quantitative

comparisons between the techniques are made, and comparisons made with results from a simple water balance ‘bucket’ model

and a Richards equation based model. It was found that the connector-type TDR sensors produced soil moisture measurements

within the ^2.5% v/v accuracy specification of the manufacturer as compared to thermogravimetric data when using the

manufacturer’s calibration relationship. However, comparisons with the water balance model showed that Virrib and buriable-

type TDR sensors yielded soil moisture changes that exceeded rainfall amounts during infiltration events. It was also found that

the CS615 reflectometer yielded physically impossible soil moisture measurements (greater than the soil porosity) during

periods of saturation. Moreover, the buriable-type TDR measurements of soil moisture content were systematically less than the

Virrib measurements by approximately 10% v/v. In addition to the good agreement with thermogravimetric measurements, the

connector-type TDR soil moisture measurements yielded the best agreement with Richards equation based model predictions of

soil moisture content, with Virrib sensors yielding a poor agreement in the deeper layers. This study suggests that connector-

type TDR sensors give the most accurate measurements of soil moisture content out of the sensor types tested.

q 2004 Published by Elsevier B.V.

Keywords: Soil moisture; In situ measurement; Techniques; Comparison

1. Introduction

It has long been recognised that reliable, robust and

automated techniques for the measurement of soil

moisture content can be extremely useful, if not

essential, in hydrologic, environmental and agricultural

0022-1694/$ - see front matter q 2004 Published by Elsevier B.V.

doi:10.1016/j.jhydrol.2004.01.008

Journal of Hydrology xx (0000) xxx–xxx

www.elsevier.com/locate/jhydrol

1 Present address: Now at School of Geography, The University

of Leeds, Leeds, United Kingdom.

* Corresponding author. Present address: Department of Civil and

Environmental, Environmental Applied Hydrology, The University

of Melbourne, Parkville, Victoria 3010, Australia. Tel.: þ61-

38344-5590; fax: þ61-38344-6215.

E-mail address: [email protected] (J.P. Walker).

HYDROL 14401—19/2/2004—13:36—SHYLAJA—94746 – MODEL 3

ARTICLE IN PRESS

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

http://www.elsevier.com/locate/jhydrol

UNCORRECTED PROOF

applications. Over the last 70 years, this recognition

has fostered the investment of a considerable amount

of ingenuity in developing such techniques. The

standard method of measuring soil moisture content is

the thermogravimetric method, which requires oven

drying of a known volume of soil at 105 8C and

determining the weight loss. This method is time

consuming and destructive to the sampled soil,

meaning that it cannot be used for repetitive

measurements at the same location. However, it is

indispensable as a standard method for calibration and

evaluation purposes.

Among the widely used automated soil moisture

measurement techniques are neutron scattering,

gamma ray attenuation, soil electrical conductivity

(including electrical conductivity probes, electrical

resistance blocks and electromagnetic induction),

tensiometry, hygrometry (including electrical resist-

ance, capacitance, piezoelectric sorption, infra-red

absorption and transmission, dimensionally varying

element, dew point, and psychometric), and soil

dielectric constant (including capacitance and time

domain reflectometry). Reviews on the advantages,

disadvantages, and basis of these measurement

techniques may be found in Wilson, (1971);

Schmugge et al. (1980); Zegelin (1996); Topp (2003).

While there is a wide range of point soil moisture

measurement techniques commercially available for

monitoring of soil moisture content, and each of the

techniques have been evaluated using the standard

thermogravimetric technique, there has been little in

the way of quantitative inter-comparison between the

various techniques, or evaluation under field con-

ditions. This paper compares the soil moisture

measurements obtained with four different soil

moisture sensors during a 2-year field study. A variety

of techniques are used in the evaluation, including

comparison of a single technique with the standard

thermogravimetric method, inter-comparisons

between the techniques, and comparisons with pre-

dictions from generally accepted models.

2. Field data and soil moisture sensors

The field data used in this study are from the

Nerrigundah experimental catchment, located in a

temperate region of eastern Australia. The main

objective of this rangeland experimental catchment

was to enable a soil moisture assimilation study at the

catchment scale. A detailed description of the

experimental catchment and the entire data set is

given in Walker et al. (2001a), so only the pertinent

details are given here.

The Nerrigundah catchment was instrumented to

monitor evapotranspiration, precipitation and soil

moisture from 12 October 1996 to 20 October 1998.

The soil moisture instrumentation, contained within a

1 £ 1 m portion of soil located in an area of negligible

lateral redistribution, consisted of Virribw soil moist-

ure sensors (the mention of trade and company names

is for the benefit of the reader and does not imply an

endorsement of the product), Soil Moisture Equip-

ment Corporation TRASEw buriable- and connector-

type time domain reflectometry (TDR) soil moisture

sensors, and a Campbell Scientific CS615 water

content reflectometer (see Fig. 1). Thermogravimetric

measurements were also made. There were no

instrument failures during the 2-year period.

2.1. Virrib sensors

The Virrib soil moisture sensors consist of two

stainless steel concentric circular rings (electrodes of

diameters 28 and 20 cm). Measurements of soil

moisture content using the Virrib sensors are made

by means of an electro-magnetic wave between these

two electrodes (Komin, Technical Data). The sensor

produces an output between 5 and 55 mA, which

corresponds to a soil moisture content range from 5 to

55% v/v. Soil moisture measurements using the Virrib

sensors are reported to be independent of the soil’s

chemical properties (Komin, Technical Data). Due to

the diameter of the outer electrode and the layer

thickness over which the sensor output responds

(approximately 12 cm when installed horizontally),

the sensor provides average soil moisture measure-

ments for a 20 l volume of soil (Komin, Technical

Data).

A total of five Virrib sensors were installed

horizontally at depths of 10, 15, 20, 30, and 40 cm to

continuously monitor soil moisture throughout the

46 cm soil profile, with measurements logged every

15 min. The minimum depth at which the Virrib sensor

could be installed without having interference from the

air layer above was 10 cm (Komin, Technical Data).


J.P. Walker et al. / Journal of Hydrology xx (0000) xxx–xxx2

ARTICLE IN PRESS

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

UNCORRECTED PROOF

Given the design of the Virrib sensors, installation

required excavation and recompaction of the soil in

which the sensors were placed for measurement of soil

moisture content. To minimise the effects of soil

disturbance, the soil was replaced in the same order

in which it was removed, with as little soil

mixing as possible. Due to the disturbance of the

soil, these sensors are reported to generally

require a few months settling time before

representative soil moisture measurements may be

made (Komin, Technical Data).

2.2. Buriable-type TDR sensors

The buriable-type TDR sensors consist of three

20 cm long waveguides and the TRASE signal unit.

By the TDR technique, measurements of soil moisture

content are made through a relationship with the

velocity of an electromagnetic wave that is passed

along the waveguides, determined by measuring the

time-of-travel. These sensors provide an average soil

moisture measurement over a layer thickness of

approximately 4 cm when installed horizontally,

with a specified accuracy of ^2.5% v/v when used

in typical mineral soils with the manufacturer’s

standard calibration relationship (Soil Moisture

Equipment Corp., 1989). While this technique can

be automated, it was not done in this application.

Periodic measurements of soil moisture using

horizontally installed buriable-type TDR sensors at

depths of 5, 10, 15, 20, 30 and 40 cm were made on a

fortnightly basis, apart from September 1997 when

they were made every 2–3 days. The minimum depth

at which the sensor could be installed without causing

a loss of accuracy was 5 cm (Soil Moisture Equipment

Corp., 1989). As the sensors consist of straight

waveguides, these sensors could be inserted into

undisturbed soil from the side of the excavation.

Therefore, measurements made using these sensors

should not be affected by disturbance to the soil from

excavation to the same extent as the Virrib sensors.

However, the disturbance caused by the actual

insertion of the waveguide into the soil may be

significant for larger waveguide diameters. Rothe et al.

(1997)) have shown that merely pushing the TDR

waveguides into the soil results in a reduction of the

measured soil moisture content of up to 10% v/v, with

the effects being strongest close to saturation. There-

fore, for waveguide diameters greater than 6 mm,

Rothe et al. (1997) suggest that it is necessary to

remove soil prior to waveguide installation by

drilling. The buriable-type TDR sensors have

Fig. 1. (a) Campbel Scientific CS615 water content reflectometer; (b) Virribw soil moisture sensor; (c) Soil Moisture Equipment Corporation

TRASEw buriable- and (d) connector-type TDR soil moisture sensor.


J.P. Walker et al. / Journal of Hydrology xx (0000) xxx–xxx 3

ARTICLE IN PRESS

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

UNCORRECTED PROOF

a waveguide diameter of 3.2 mm. However, due to the

dry state of the soil at the time of installation, the soil

was extremely hard and installation was difficult,

requiring pilot holes to be formed prior to installation

of the sensors. Thus the installation of these types of

sensors could be prone to suffer from gaps between

the soil and the waveguide.

2.3. Connector-type TDR sensors

The connector-type TDR sensors consist of two

stainless steel waveguides of user specified length, a

balun and the TRASE signal unit. These sensors are

reported to have the same area of influence and

accuracy specifications as the buriable-type sensors.

The connector-type TDR sensors consist of two 6 mm

diameter waveguides that are inserted from the soil

surface. Given the design and diameter of these

waveguides, they could be inserted from the surface

into undisturbed soil without pre-forming holes, even

under relatively dry soil conditions.

The vertically inserted connector-type TDR sen-

sors provided an average soil moisture measurement

over depths of 0–5, 0–10, 0–15, 0–20, 0–30, and

0–40 cm, being the length of the waveguides used.

While this technique can be automated, in this

application measurements were made on a fortnightly

basis, apart from September 1997 when they were

made every 2–3 days. These sensors were not

installed until 24 April 1997.

There are upper and lower limitations on the length

of TDR waveguides that may be used. The upper limit

on waveguide length is governed by the strength-of-

arm of the person who inserts the waveguide.

However, a more severe limit to waveguide length

arises from loss of TDR signal in the soil (Zegelin,

1996). To overcome the arm strength limitations,

waveguides of greater than 15 cm length were

inserted by hammering employing a waveguide

insertion tool. The lower limit on waveguide length

is imposed by the accuracy of the time-of-travel

measurement of the TDR device, which is currently of

order 0.1 ns, limiting waveguide length to greater than

5 cm (Zegelin, 1996). However, Zegelin (1996) has

noted that waveguide lengths of 10 cm even have a

reduced accuracy because of this timing limit, and

Soil Moisture Equipment Corp., 1989) warns against

using waveguide lengths of less than 15 cm due to

a loss of accuracy. The impact of shorter waveguides

on measurement accuracy is discussed further in the

section on thermogravimetric measurements.

2.4. Campbell scientific CS615 sensors

The Campbell Scientific CS615 water content

reflectometer consists of two 30 cm long stainless

steel waveguides connected to a printed circuit board,

and measures the soil moisture content using the TDR

technique (Campbell Scientific Inc., 1995). The

CS615 reflectometer is specified to have an accuracy

of ^2.5% v/v when applied to typical mineral soils

using the manufacturer’s standard calibration

relationship. Soils with different dielectric properties

show an error that appears as a constant offset

(Campbell Scientific Inc., 1995). However, both the

accuracy and stability of the sensor are affected by the

soil’s electrical conductivity. With soil electrical

conductivity above 2 dSm21 the sensor output

changes and at electrical conductivity values greater

than 20 dSm 2 1 the sensor output becomes unstable

(Campbell Scientific Inc., 1995). An important

consideration with the CS615 reflectometer is its

strong dependence on soil temperature. To account for

this temperature dependence, a temperature correc-

tion polynomial has been supplied (Campbell Scien-

tific Inc., 1995).

A single CS615 sensor was installed horizontally

into undisturbed soil from the side of an excavation at

a depth of 5 cm, providing a soil moisture measure-

ment over a layer thickness of approximately 4 cm

(Campbell Scientific Inc., 1995). Thermocouples

were installed at depths of 4 and 6 cm to enable the

temperature correction to be made. These measure-

ments were logged every 10 min from 8 May 1997.

The CS615 reflectometer has a waveguide diameter of

3.2 mm (the same as the buriable-type TDR sensors),

but due to the moist state of the soil at the time of

installation, the sensor could be easily installed

without pre-forming holes.

2.5. Thermogravimetric measurements: sensor

calibration

An in-situ calibration of the soil moisture sensor

installations described above could not be performed

without destroying the soil moisture monitoring site.



ARTICLE IN PRESS

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

UNCORRECTED PROOF

Hence, evaluation of soil moisture measurements was

performed by making comparisons between the

different soil moisture sensor types, and using a

calibration of the connector-type TDR sensor to give

confidence in those measurements. The calibration of

5, 10 and 15 cm waveguide lengths was evaluated from

thermogravimetric field samples (10 cm diameter soil

sampling ring of 5 cm depth). The calibration of longer

waveguides was not evaluated due to the destructive

nature and labour intensiveness of the testing, and the

number of calibration data values required to make

conclusive statements regarding accuracy, and the

good agreement for the shorter waveguides. In

addition, literature suggests that longer waveguides

should not result in any further loss of accuracy for the

waveguide lengths used. Hence, provided satisfactory

calibration results were obtained for the shorter

waveguides, measurements made with longer wave-

guides should also be of sufficient accuracy.

The comparison of connector-type TDR and

thermogravimetric measurements is given in Fig. 2,

where it can be seen that measurements from the 10

and 15 cm waveguides are in good agreement with

thermogravimetric measurements when using the

manufacturer’s calibration relationship with approxi-

mately the ^2.5% v/v accuracy stated by the

manufacturer. However, the comparison with 5 cm

waveguides may be interpreted in one of three ways:

(i) TDR soil moisture measurements follow a 1:1

relationship with the thermogravimetric measure-

ments but have a very low accuracy, approxi-

mately ^7% v/v. This interpretation requires the

assumption that the lack of spread around the 1:1

line at soil moisture contents below 40% v/v is due

to an insufficiently large sample size.

(ii) There is a non-linear or non-continuous relation-

ship between TDR and thermogravimetric soil

moisture measurements. Using this interpret-

ation, individual relationships may be fitted to

TDR soil moisture measurements above and

below 25% v/v.

(iii) For soil moisture content less than 15% v/v the

TDR method cannot measure soil moisture

content reliably using 5 cm waveguides, with a

variation of 20% v/v in the TDR measurements

for the same thermogravimetric soil moisture

content.

Fig. 2. Comparison of thermogravimetric and connector-type TDR

soil moisture measurements for varying waveguide lengths: (a)

5 cm; (b) 10 cm; and (c) 15 cm.



ARTICLE IN PRESS

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

UNCORRECTED PROOF

A reason to support (iii) is that as soil becomes

drier, the dielectric constant of the soil is reduced,

and hence the velocity of the electromagnetic wave

is increased. The effect of this increase in velocity

would be to make determination of the travel time

along the waveguide more difficult due to the

shortness of the waveguide. Timing errors will also

have a greater influence on the soil moisture content

measurement with shorter waveguides. From this

third interpretation, a linear relationship may be

fitted to TDR measurements above 25% v/v, and

any TDR soil moisture measurements below 25%

v/v regarded as erroneous. Because of the uncer-

tainty associated with 5 cm waveguide measure-

ments, this data is only used in the qualitative

comparison section of this paper.

3. Qualitative comparison of techniques: sensor

intercomparison

An intercomparison of soil moisture as measured

by the various electronic techniques described above

is given in Fig. 3 for depth variation at discrete times,

and Fig. 4 for temporal variation in the near-surface

measurements. Fig. 5a shows an intercomparison of

the temporal variation of soil moisture in the top

40 cm of the soil profile.

The intercomparisons in Fig. 3 show that the

Virrib sensors continually gave soil moisture

measurements that were approximately 10% v/v

higher than the buriable-type TDR sensors, while

the connector-type TDR sensors gave mid-range soil

moisture contents. Moreover, plotting of the soil

moisture profile measurements at discrete times

revealed that the soil moisture measurements did

not yield a smooth variation of soil moisture content

with depth, and that disaggregation of the con-

nector-type TDR measurements for the soil moisture

profile was sometimes difficult (e.g. 1 October), as

small differences in soil moisture measurements

yielded large differences in the layer estimates when

performing the disaggregation. Thus, aggregated

rather than disaggregated comparisons are made in

the remainder of this paper. The non-smooth

variation of soil moisture with depth for Virrib

and buriable-type TDR may be a result of: (i)

inaccurate measurement of soil moisture content by

Fig. 3. Depth comparison of Virrib, CS615, buriable- and

connector-type TDR soil moisture sensor measurements for various

dates.



ARTICLE IN PRESS

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

UNCORRECTED PROOF

the sensors; or (ii) physical differences in soil

moisture content with depth as a result of natural

variation in soil properties. The latter is more likely.

The inset of Fig. 4 shows that both the Virrib and

CS615 reflectometer have diurnal variations in the

soil moisture measurements. While the CS615

reflectometer measurements do not exhibit the high

frequency noise present in the Virrib measurements,

which requires filtering before the data can be used,

soil moisture measurements as high as 70% v/v were

recorded. Such high soil moisture measurements were

physically unrealistic for this soil, with a porosity of

approximately 60% v/v. The CS615 reflectometer

measurements indicated higher soil moisture content

Fig. 4. Temporal comparison of soil moisture measurements: Virrib sensor at 10 cm (solid line), CS615 sensor at 5 cm (dashed line), buriable-

type TDR sensor at 5 cm (solid square) and 10 cm (open square), and connector-type TDR for 0-5 cm (solid circle) and 0–10 cm (open circle).



ARTICLE IN PRESS

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

UNCORRECTED PROOF

than Virrib measurements during wet periods and

lower soil moisture content during dry periods. This is

to be expected, with the CS615 reflectometer installed

at a shallower depth. Moreover, the same overall trend

was apparent and there was a reasonable agreement

between CS615 reflectometer and buriable-type TDR

measurements, particularly during the southern hemi-

sphere summer of 1997/98.

It is difficult to draw any general conclusions

regarding the comparison of connector-type TDR

measurements with the other instrument types pre-

sented in Fig. 4, or the overall accuracy of any

particular system from Figs. 3 or 4. However, Fig. 5(a)

shows a generally good agreement between the Virrib

and connector-type TDR measurements of soil

moisture content in the top 40 cm of the soil profile

and a poorer agreement with buriable-type TDR

measurements. Qualitatively similar results were

obtained for comparisons of aggregated soil moisture

over shallower depths. This would suggest that Virrib

and connector-type TDR measurement techniques

were the most accurate of the four techniques tested,

based on the qualitatively similar buriable-type TDR

and CS615 reflectometer soil moisture measurements

at 5 cm depth, and the good agreement of connector-

type TDR and thermogravimetric measurements.

4. Quantitative comparison of techniques: soil

considerations

An investigation into the variability of soil

moisture content over short length scales was

undertaken to see if the differences between Virrib,

buriable- and connector-type TDR soil moisture

measurements were due to actual differences in soil

Fig. 5. (a) Comparison of soil moisture measurements for Virrib (solid circle), buriable- (open circle with dot) and connector-type TDR (open

circle), to a soil depth of 40 cm; (b) Comparison of cumulative change in soil moisture profile storage for Virrib, buriable- and connector-type

TDR, and a bucket water balance model (open square).



ARTICLE IN PRESS

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

742

743

744

745

746

747

748

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765

766

767

768

UNCORRECTED PROOF

moisture as a result of natural variation in soil

properties. For this investigation, transects of soil

moisture were measured with the 15 cm connector-

type TDR waveguides every 0.5 m for a distance of

25 m and every 0.1 m for a distance of 2.5 m, under

saturated and unsaturated conditions. Measurements

were made in a level area near the intensive soil

moisture instrumentation site, with variograms and

autocorrelation relationships determined for each of

the data sets.

Under saturated soil conditions the measurements

reflect the natural variation in soil porosity, whilst the

other measurements reflect the variation in soil

moisture content due to natural variations in

soil properties (such as soil texture), and noise in the

measurement technique. The results from this analysis

indicated a very short correlation length (less than

0.5 m), with the variogram suggesting a nugget effect

due to error in the TDR measurement technique of

between 1 and 3% v/v, which is consistent with the

manufacturer’s specification.

As the large difference between Virrib and

buriable-type TDR soil moisture measurements

could not be explained by a short scale natural

variation in the soil properties, an alternative

explanation was sought. The soil disturbance caused

by installation of the Virrib sensors was considered to

be a contributing factor. Therefore, measurements

were made using 15 and 30 cm connector-type TDR

waveguide lengths in the disturbed soil where the

Virrib sensors were installed, and in the undisturbed

soil near the buriable-type TDR sensors. The soil

moisture measurements in the disturbed soil (31 July

1997) were 38.2 and 37.4% v/v for the 15 and 30 cm

waveguides, respectively, whilst the soil moisture

measurements in the undisturbed soil were 35.9 and

35.6% v/v for the 15 and 30 cm waveguides,

respectively. Soil moisture measurements were

approximately 2% v/v drier in the undisturbed soil

in both instances.

This analysis suggests that some of the difference

between Virrib and buriable-type TDR soil moisture

measurements is attributable to soil disturbance

during installation of the Virrib sensors, even nine

months after installation. However, this does not

account for the entire difference observed, which is

possibly due to gaps between the soil and the buriable-

type TDR waveguides. Any air or fluid filled gaps

around the waveguide due to insertion affect the

ability of the TDR technique to measure the soil

moisture content accurately. The effect of gaps is

reported to be greater for three-rod sensors (i.e.

buriable-type TDR) than two-rod sensors (i.e. con-

nector-type TDR and CS615 reflectometer), and if

the gaps are filled with water rather than air (Knight

et al., 1997).

5. Quantitative comparison of techniques: simple

analysis

Immediately following periods of infiltration (14

February 1997, 22 May 1997 and 28 April 1998), the

soil moisture measurements by the Virrib sensors

indicated systematic increases in soil water storage

greater than those indicated by both the buriable- and

connector-type TDR soil moisture measurements (Fig.

5a). To identify if the Virrib sensors were over-

estimating or if the TDR sensors were under-estimat-

ing the changes in soil moisture content, a comparison

of cumulative change in soil moisture storage based on

the soil moisture measurements was made with a

simple bucket type water balance model (Fig. 5b).

In the simple bucket model, it was assumed there

was no drainage from the bottom of the soil profile, all

rainfall infiltrated up to the maximum soil moisture

storage (porosity £ soil depth), and actual evapotran-

spiration was estimated from the Penman-Monteith

potential evapotranspiration, reduced by a soil

moisture stress index. The soil moisture stress index

used in this study was the average column soil

moisture content divided by the average column

porosity. There was no calibration of the model. Soil

moisture storage calculations were commenced from

installation of the connector-type TDR probes (24

April 1997), and were normalised so that soil moisture

storage estimates were the same for each sensor type

at commencement of calculations. This removed the

systematic bias between Virrib and buriable-type

TDR soil moisture measurements. The results from

this analysis indicate that the Virrib sensors were

over-estimating the changes in soil moisture content,

with connector-type TDR soil moisture measurements

and the crude water balance calculations having a

good agreement for the two major infiltration events

when there was data from all three sensor types.



ARTICLE IN PRESS

769

770

771

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

796

797

798

799

800

801

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

817

818

819

820

821

822

823

824

825

826

827

828

829

830

831

832

833

834

835

836

837

838

839

840

841

842

843

844

845

846

847

848

849

850

851

852

853

854

855

856

857

858

859

860

861

862

863

864

UNCORRECTED PROOF

Analysis of the buriable-type TDR measurements

indicated close agreement with the changes in soil

water storage from the Virrib sensors.

The other obvious discrepancy between Virrib and

connector-type TDR soil moisture measurements

(Fig. 5a) was the dry down period from July 1997 to

October 1997. During this period, data from the Virrib

sensor showed a consistently wetter estimate of the

soil moisture content. The comparison with water

balance calculations (Fig. 5b) was again best for the

connector-type TDR measurements. The discrepancy

of water balance calculations with measurements

from October 1997 to January 1998 was a result of the

model assumption that all rainfall infiltrated. At lower

soil moisture contents and for heavy rainfall this is not

the case, with infiltration capacities being less than the

rainfall rate. Due to the poor agreement between

connector- and buriable-type TDR data, and the

concern over the installation of buriable-type TDR

sensors resulting in gaps, the buriable-type TDR

measurements are not used further in this paper.

6. Quantitative comparison of techniques: detailed

analysis

This section uses a one-dimensional Richards

equation based soil moisture model (see Walker

et al., 2001b for details) to further explore the

accuracy of Virrib and connector-type TDR soil

moisture data. To allow instrument settling time and

avoid the period of erroneous Virrib soil moisture data

identified above (moisture data suggested that more

rainfall occurred than was recorded by the raingauges

during wetting up periods), the soil moisture model

was calibrated for the 100 day dry down period from

16 June 1997 (Julian day 167) to 24 September 1997

(Julian day 267); approximately eight months after

installation of the Virrib sensors.

6.1. Observed model parameters

Several of the parameters required by the soil

moisture model could be defined directly from field

observations and measurements. Soil layer thicknesses

were set to the observed soil horizon thicknesses with

the exception of the surface layer, which was set to a

thickness of 1 cm. The depression storage parameter

was set at 5 mm, based on measurements of root mean

square surface roughness. Likewise, the saturated

hydraulic conductivity ðKSÞ was estimated from

Guelph permeameter and double ring infiltrometer

measurements. The porosity ðfÞ and residual soil

moisture content ðurÞ for each of the model layers were

estimated from an analysis of the soil moisture

measurements over the entire record of data. These

parameter values are given in Tables 1 and 2.

6.2. Calibrated model parameters

The model parameters requiring calibration were

the van Genuchten parameter n (van Genuchten,

1980), for relating hydraulic conductivity to saturated

hydraulic conductivity, and MGRAD, a maximum

matric head gradient parameter (Walker et al., 2001b)

that is dependent on soil type and used to estimate the

matric head gradient for vertical redistribution of soil

moisture. Calibration of these parameters was per-

formed with the Bayesian non-linear regression

program NLFIT. The NLFIT program suite (Kuczera,

1994) is an interactive optimisation package, employ-

ing the Shuffled Complex Evolution Method

Table 1

Soil parameters for the one-dimensional soil moisture model from calibration to Virrib soil moisture data

Layer Thickness (mm) Horizon Observed Calibrated

KS (mm/h) f (% v/v) ur (% v/v) n MGRAD (mm)

1 10 A1 15 60 6 1.15 432

2 45 A1 15 60 6 1.15 432

3 68 A2 15 46 8 2.42 384

4 112 B1 3 42 12 1.64 389

5 225 B2 0.4 48 18 2.25 32



ARTICLE IN PRESS

865

866

867

868

869

870

871

872

873

874

875

876

877

878

879

880

881

882

883

884

885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903

904

905

906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929

930

931

932

933

934

935

936

937

938

939

940

941

942

943

944

945

946

947

948

949

950

951

952

953

954

955

956

957

958

959

960

UNCORRECTED PROOFdeveloped by Duan et al. (1994). The data used for

calibration were the Virrib soil moisture measure-

ments of model layers 3 and 4 (model layers 1 and 2

were too shallow for comparison with soil moisture

measurements and soil moisture measurements of

model layer 5 were not used as a good fit to the data

could not be obtained when using those measure-

ments) and the connector-type TDR measurements of

depth integrated soil moisture over model layers 1–5

(see Table 3 for depth comparisons). The n and

MGRAD parameter values from calibration to the

different data types are given in Tables 1 and 2.

Initial soil moisture values were set to the soil

porosity values, given that calibration commenced

when the soil was saturated. With the soil column

being underlain by a low permeability sandstone, a

zero-moisture flux boundary condition was applied to

the base of the soil column. The surface soil moisture

flux boundary condition was set at a fixed value for

half hour simulation periods. This surface soil

moisture flux was taken as the average of 10 min

measurements of Penman–Monteith potential evapo-

transpiration rate, reduced by the soil stress index,

except for periods during which rainfall was recorded.

During these periods, it was assumed that no

evapotranspiration occurred and that the rainfall

recorded had a uniform rainfall rate over the half

hour period. Ponding greater than the depression

storage depth was instantly disposed of as runoff.

6.3. Calibration to virrib measurements

Parameter values from calibrating to the Virrib

observations (Table 1) gave a very good agreement

between the model predictions and Virrib soil

moisture data (Fig. 6) in model layers 3 and 4

(comparison with layer 4 not shown). However, the

comparison with Virrib soil moisture data in model

layer 5 was very poor from Julian day 200–260. It

was believed that this poor agreement in model layer 5

was a result of poor quality soil moisture data in this

layer, rather than a weakness of the model. Installation

of the Virrib soil moisture sensors required excavation

and recompaction of the soil around the sensors.

Hence, even after eight months the soil may not have

returned to its original state, particularly at greater

depths. In addition, just prior to the calibration period

the Virrib soil moisture sensors over-responded to the

increase in soil moisture content relative to the

amount of rainfall recorded. It was also found that

Table 2

Soil parameters for the one-dimensional soil moisture model from calibration to connector-type TDR soil moisture data (from Walker et al.,

2001b)

Layer Thickness (mm) Horizon Observed Calibrated

KS (mm/h) f (% v/v) ur (% v/v) N MGRAD (mm)

1 10 A1 15 60 6 2.18 23

2 45 A1 15 60 6 2.18 23

3 68 A2 15 46 8 1.34 22

4 112 B1 3 42 12 2.19 50

5 225 B2 0.4 48 18 1.46 275

Table 3

Model and measurement depths used in comparisons

Model Virrib Model Connector

Layers Depth (mm) Sensors Depth (mm) Layers Depth (mm) Waveguide length(s) (mm) Depth (mm)

3 55–123 1 40–160 1–3 0–123 100/150 0–125

4 123–235 2/3 90–260 1–4 0–235 200/300 0–250

5 243–460 4/5 260–460 1–5 0–460 400 0–400



ARTICLE IN PRESS

961

962

963

964

965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987

988

989

990

991

992

993

994

995

996

997

998

999

1000

1001

1002

1003

1004

1005

1006

1007

1008

1009

1010

1011

1012

1013

1014

1015

1016

1017

1018

1019

1020

1021

1022

1023

1024

1025

1026

1027

1028

1029

1030

1031

1032

1033

1034

1035

1036

1037

1038

1039

1040

1041

1042

1043

1044

1045

1046

1047

1048

1049

1050

1051

1052

1053

1054

1055

1056

UNCORRECTED PROOF

connector TDR soil moisture measurements did not

agree well with the Virrib soil moisture measurements

during this period, and that connector TDR soil

moisture measurements agreed better with soil

moisture calculations using a simple bucket water

balance model (Fig. 5b).

A comparison of connector-type TDR soil moist-

ure measurements with model predictions from this

Fig. 6. Comparison of Virrib soil moisture measurements (open circles) with the Richards equation based soil moisture model prediction (solid

line): (a) model layer 3 and (b) model layer 5 with calibration to Virrib soil moisture measurements; (c) model layer 3 and (d) model layer 5 with

calibration to connector-type TDR soil moisture measurements.



ARTICLE IN PRESS

1057

1058

1059

1060

1061

1062

1063

1064

1065

1066

1067

1068

1069

1070

1071

1072

1073

1074

1075

1076

1077

1078

1079

1080

1081

1082

1083

1084

1085

1086

1087

1088

1089

1090

1091

1092

1093

1094

1095

1096

1097

1098

1099

1100

1101

1102

1103

1104

1105

1106

1107

1108

1109

1110

1111

1112

1113

1114

1115

1116

1117

1118

1119

1120

1121

1122

1123

1124

1125

1126

1127

1128

1129

1130

1131

1132

1133

1134

1135

1136

1137

1138

1139

1140

1141

1142

1143

1144

1145

1146

1147

1148

1149

1150

1151

1152

UNCORRECTED PROOF

calibration (Fig. 7) found a good agreement for the

average soil moisture content of model layers 1–4

(comparison not shown) and model layers 1–5, but a

rather poor agreement for the average soil moisture

content of model layers 1–3. Furthermore, the model

simulation did not respond to the wetting up towards

the end of the calibration period, as indicated by the

observed data.

Fig. 7. Comparison of connector-type TDR soil moisture measurements (open circles) with the Richards equation based soil moisture model

prediction (solid line): (a) model layers 1 to 3 and (b) model layers 1–5 with calibration to Virrib soil moisture measurements; (c) model layers

1–3 and (b) model layers 1–5 with calibration to connector-type TDR soil moisture measurements.



ARTICLE IN PRESS

1153

1154

1155

1156

1157

1158

1159

1160

1161

1162

1163

1164

1165

1166

1167

1168

1169

1170

1171

1172

1173

1174

1175

1176

1177

1178

1179

1180

1181

1182

1183

1184

1185

1186

1187

1188

1189

1190

1191

1192

1193

1194

1195

1196

1197

1198

1199

1200

1201

1202

1203

1204

1205

1206

1207

1208

1209

1210

1211

1212

1213

1214

1215

1216

1217

1218

1219

1220

1221

1222

1223

1224

1225

1226

1227

1228

1229

1230

1231

1232

1233

1234

1235

1236

1237

1238

1239

1240

1241

1242

1243

1244

1245

1246

1247

1248

UNCORRECTED PROOF

6.4. Calibration to connector TDR measurements

To improve the comparison of model predictions

with the observed connector-type TDR soil moisture

measurements, a calibration was made to the

connector-type TDR soil moisture data. The par-

ameter values from this calibration (Table 2) gave a

very good agreement between model predictions and

the connector-type TDR soil moisture data (Fig. 7) for

model layers 1–3 and 1–4 (comparison not shown),

and an equally good agreement for model layers 1–5

as for the calibration to Virrib soil moisture data. The

model simulation also responded to the wetting up at

the end of the calibration period. In addition, the

MGRAD values from calibration to the connector-

type TDR data increased with depth, rather than

decrease as with the calibration to Virrib soil moisture

data. Increasing MGRAD values with depth agree

better with intuition, as the clay content in the soil

profile increased with depth, and hence the maximum

matric suction gradient should be greater at deeper

depths. The consistent slight over-estimation of soil

moisture storage for the entire soil profile as compared

to the connector-type TDR measurements may be a

result of: (i) the no drainage boundary condition at the

bottom of the soil profile, (ii) an incorrect estimate of

rainfall and/or evapotranspiration, or (iii) a systematic

error in the 40 cm TDR measurement. A joint

calibration to Virrib and connector-type TDR data

and/or use of a gravity drainage boundary condition

was unable to improve the results.

A comparison of the model prediction is made with

the Virrib soil moisture data in Fig. 6. This

comparison shows a good agreement with model

layer 3 and an even better agreement in model layer 5

than for the calibration to Virrib soil moisture data

itself. However, the comparison with model layer 4

(comparison not shown) was slightly degraded. This

analysis provides further support to the accuracy of

connector-type TDR soil moisture measurements over

the Virrib (and other) techniques tested in this paper.

7. Conclusions

This paper found a good agreement between the

Soil Moisture Equipment Corporation TRASEw

connector-type time domain reflectometry (TDR)

and thermogravimetric soil moisture measurements.

However, there were significant differences between

the other electronic soil moisture techniques studied,

which included Virribw soil moisture sensors, a

Campbell Scientific CS615 water content reflect-

ometer, and Soil Moisture Equipment Corporation

TRASE buriable-type TDR soil moisture sensors. The

conclusion drawn from an analysis of these measure-

ments was that the differences in the soil moisture

measurements from the various sensors were the

result of a combination of factors.

Firstly, the installation procedure for the Virrib

sensors involved major disturbance to the soil in

which soil moisture was being measured, thus altering

the physical properties of the soil, and still influencing

the physical moisture content of the soil in compari-

son to the undisturbed soil even nine months after

installation. Secondly, the different sensors use

different measurement techniques and measure the

soil moisture of different size volumes of soil. Thirdly,

any air or fluid filled gaps around the TDR

waveguides due to insertion affect the ability of the

TDR technique to measure the moisture content of the

soil accurately, with the effects being greater for

buriable-type sensors (having three-prong wave-

guides) than the connector-type and CS615 reflect-

ometer sensors (having two-prong waveguides). Thus

the dry soil conditions at time of installation for the

buriable-type TDR sensors may have resulted in poor

installation of the buriable-type TDR sensors, result-

ing in gaps which introduce further errors in the

measurements, especially under wet soil conditions.

Both the Virrib and CS615 reflectometer soil

moisture sensors displayed diurnal variations in soil

moisture content and the Virrib sensors displayed high

frequency noise that required filtering. Moreover,

comparison with a simple water balance ‘bucket’

model showed that measured changes in soil moisture

storage using the Virrib and buriable-type TDR sensors

exceeded the recorded rainfall amounts during infiltra-

tion events. Furthermore, the CS615 reflectometer soil

moisture sensor predicted soil moisture contents that

exceeded the soil porosity during periods of saturation.

While buriable-type TDR sensors yielded system-

atically lower measurements of soil moisture content

in comparison to Virrib measurements, these two

sensor types yielded similar estimates of change

in water storage. Comparison between Virrib and



ARTICLE IN PRESS

1249

1250

1251

1252

1253

1254

1255

1256

1257

1258

1259

1260

1261

1262

1263

1264

1265

1266

1267

1268

1269

1270

1271

1272

1273

1274

1275

1276

1277

1278

1279

1280

1281

1282

1283

1284

1285

1286

1287

1288

1289

1290

1291

1292

1293

1294

1295

1296

1297

1298

1299

1300

1301

1302

1303

1304

1305

1306

1307

1308

1309

1310

1311

1312

1313

1314

1315

1316

1317

1318

1319

1320

1321

1322

1323

1324

1325

1326

1327

1328

1329

1330

1331

1332

1333

1334

1335

1336

1337

1338

1339

1340

1341

1342

1343

1344

UNCORRECTED PROOF

connector-type TDR measurements of total profile

water storage was generally good, with the exception

of the major infiltration events and towards the end of

the major dry down event. In addition to the good

agreement with thermogravimetric measurements, the

connector-type TDR soil moisture measurements

yielded better agreement with Richards equation

based model predictions of soil moisture than did the

Virrib measurements, providing further evidence that

these sensors yielded the most accurate measurements

of soil moisture content. It was not possible to attain a

good agreement between Richards equation based

model predictions and Virrib measurements of deep

soil moisture content, even with calibration to that

data.

Acknowledgements

This work has been jointly funded by an Australian

Postgraduate Award scholarship and the Hunter Water

Corporation. Andrew Krause and Scott Wooldridge

are acknowledged for their assistance with field work.

John and Margaret Russell are acknowledged for the

use of their land in the monitoring of soil moisture.

References

Campbell Scientific Inc.,, 1995. Water Content Reflectometer

InstructionManual,CampbellScientificInc.,Logan,USA,p.8pp.

Duan, Q., Sorooshian, S., Gupta, V.K., 1994. Optimal use of the

SCE-UA global optimization method for calibrating watershed

models. J. Hydrol. 158, 265–284.

Knight, J.H., Ferre, P.A., Rudolph, D.L., Kachanoski, R.G.,

1997. A Numerical analysis of the effects of coatings and

gaps upon relative dielectric permittivity measurement with

time domain reflectometry. Water Resour. Res. 33 (6),

1455–1460.

Kuczera, G., 1994. NLFIT: a Bayesian Non-linear Regression

Program Suite. Department of Civil Engineering and Surveying,

The University of Newcastle, Australia.

Rothe, A., Weis, W., Kreutzer, K., Matthies, D., Hess, U., Ansorge,

B., 1997. Changes in soil structure caused by the installation of

time domain reflectometry probes and their influence on the

measurement of soil moisture. Water Resour. Res. 33 (7),

1585–1593.

Schmugge, T.J., Jackson, T.J., McKim, H.L., 1980. Survey of

methods for soil moisture determination. Water Resour. Res. 16

(6), 961–979.

Soil Moisture Equipment Corp., 1989. Trase System 1 Operating

Instructions. Soil Moisture Equipment Corp., 54.

Topp, G.C., 2003. State of the art of measuring soil water content.

Hydrol. Process. 17, 2993–2996.

van Genuchten, M.T., 1980. A closed-form equation for predicting

the hydraulic conductivity of unsaturated soils. Soil Sci. Soc.

Am. J. 44, 892–898.

Walker, J.P., Willgoose, G.R., Kalma, J.D., 2001a. The nerrigundah

data set: soil moisture patterns, Soil charachteristics and

hydrological flux measurements. Water Resour. Res. 37 (11),

2653–2658.

Walker, J.P., Willgoose, G.R., Kalma, J.D., 2001b. One dimen-

sional soil moisture profile retrieval by assimilation of near-

surface mesaurements: a simplified soil moisture model and

field application. J. Hydromet. 2, 356–373.

Wilson, R. G., 1971. Methods of Measuring Soil Moisture.

Technical Manual Series, The Secretariat, Canadian National

Commission for the International Hydrological Decade, Ottawa,

Canada, 20 pp.

Zegelin, S., 1996. Soil Moisture Measurement, Field Measurement

Techniques in Hydrology-Workshop Notes, Corpus Christi

College, Clayton, pp.C1–C22.



ARTICLE IN PRESS

1345

1346

1347

1348

1349

1350

1351

1352

1353

1354

1355

1356

1357

1358

1359

1360

1361

1362

1363

1364

1365

1366

1367

1368

1369

1370

1371

1372

1373

1374

1375

1376

1377

1378

1379

1380

1381

1382

1383

1384

1385

1386

1387

1388

1389

1390

1391

1392

1393

1394

1395

1396

1397

1398

1399

1400

1401

1402

1403

1404

1405

1406

1407

1408

1409

1410

1411

1412

1413

1414

1415

1416

1417

1418

1419

1420

1421

1422

1423

1424

1425

1426

1427

1428

1429

1430

1431

1432

1433

1434

1435

1436

1437

1438

1439

1440

Documents

In situ measurement of soil moisture: a comparison of …jpwalker/papers/jh04.pdfJeffrey P. Walker*, Garry R. Willgoose1, Jetse D. Kalma Department of Civil, Surveying and Environmental