Estimating depths to claypansusing electromagneticinduction methodsJ. A. Doolittle, K. A. Sudduth, N. R Kitchen, and S. J. Indorante

ABSTRACT: Claypans restrict infiltration, influence the lateral movement of soil water andagrichemicals. and limit cropproduction. In many areasof the Midwest, quantification andmapping of the variations in the depth to claypansare important componentsjOr researchonwater-quality and cropproduction. In an area of Mexico soils in central Missouri, a highcorrelation between the observed depth to claypans and the responseof the EM38 meter wasobserved. Equations were developed to infer depths and chart the topography of the claypan.Compared with traditional methods of observing this subsurface laye/; EM techniques arenoninvasive. less labor intensive. more economical and can produce large quantities of datain a relativelyshort period of time.

Elecrromagnetic induction (EM) tech-niques use elecrromagnetic energy to

measure the apparenr conductivity ofearthen materials. Apparenr conductiviryis the weighted average conductivitymeasurement for a column of earthen

materials to a specified observationdepth (7). The averages are weighted ac-cording to the depth response functionof the meter (16). This technology hasbeen used by geologists and geophysiciststo map glacial deposits, bedrock surfaces(20) and permafrost (9), estimate thethickness of clay deposits (I 3), locatesand and gravel deposits (11), predictsoil water content (8), and for ground-water investigations (2, 3, 7). More re-cently, this technology has been used insoil science principally to identify, map,and monitor soil saliniry and groundwa-ter contamination (4,5,6,14,16,19). Inaddition, this technology has been usedto identif}, sodium-affected soils (1) andto assess edaphic properties important toforest site productiviry (10). These stUd-ies have documented that this noninva-

sive technique is facile, provides a largenumber of measurements for the inter-

pretation of site conditions, and can be

lA. Doolirtle is a soil specialist with the USDA-Soil COl/servation Service. Chester, Pennsylvania,

19013; K. A. Sudduth is an agricultural engineerwith the Cropping Systems & Water Quality Re-search Ullit, USDA-Agriculmral Research Service,Columbia, 1\10, 65211; HR. Kitchen is a research

assistal1t professor with the School of Natural Re-sources. UIli!'ersity of Missouri, Columbia, Mis-souri, 65211; and S.j. lndoranre is a soil surveyproject leader with the USDA-Soil Conservation

Service, .\!LRA Soil Survey Office, Belleville, Illi-nois. 62220.

Trade /lames are used in this article to providespecific h!fOrmation and do not constitute endorse-mel1t ~J' the aurhors.

/. Soil filld \rimr Com. 49 (6) 512-515

applied over broad areas and soil types.Variations in electromagnetic response

are produced by changes in the ionic con-centration of earthen materials. Factors

influencing the ionic concentration andconductivity of earthen materials include(i) moisture content, (ii) amount and rypeof ions in the S9il water, (iii) and amountand type of clays in the soil matrix. Inareas of salt-affected soils, it has been esti-mated that 65% to 70% of the electro-

magnetic response can be explained bythe concentration of soluble salts (19). Innonsaline soils of humid areas, however,soil textUre, moistUre content, and cationexchange capaciry are often the principalfactors determining apparem conductiviry

(8.10). The appareIH coI1JlIcri\'ir~' of soilsincreases with exchange capacit~,. moisrurecontent, and clay content (8,15).

Generally, EM techniques have beenused most successfully in areas where sub-surface properties are reasonably homoge-neous, the effects of one facror (clay,moistUre, or salt content) dominates overthe other factors, and variations in EM

response can be related to changes in rhedominam factor (3). In such areas infor-mation is generally gathered on rhe dom-inant factor, and assumptions are madeconcerning rhe behavior of the other fac-tors (2).

In the midwestern United States. there

are aboUt 4 million tilled hectares of clay-pan soils (17). These soils exhibit a dense,fine-textured subsoil, which restricts infil-tration and influences the lateral move-

ment of soil water and agrichemicals. Inmany areas of the midwest, quantificationand mapping of the depth to claypan areimportant components for research onwater-quality and crop production. Col-lection of such data with a soil probe orauger is highly tedious and labor inten-sive, making it impractical to collect thelarge quantities of data necessary to im-plement comprehensive claypan mappingstUdies over large areas. This stUdy wasundertaken to compare the relationshipberween data on depth to claypan as in-ferred by EM methods with data collectedby conventional sampling methods. Todemonstrate the use and compatibility of

572 l()l'R~AI. OF SOil A~[} WATER COr--:SERVATION

Figure 1. Using the EM38 meter to measure apparent conductivity

EM and computer-graphic techniques,EM responses and a developed regressionequation were used co predict and mapvariations in the depths co claypan in an

adjoiningsire.The useof rhesetech-niques for mapping the depth co a strong-ly contrasting soil horizon has not beenpreviously reponed.

Materials and methods

Equipment. A Geonics Limited, EM38ground conductivity meter was used inthis srudy. This meter is shown in Figure1. Principles of operation have been de-scribed in detail by McNeill (I1,12).

The observation depths of an EM meterare dependent upon intercoil spacing,transmission frequency, and coil orienta-tion relative co the ground surface. TheEM38 meter has a fixed intercoil spacingof aboUt one meter. Ie operates at a fre-quency of 13.2 kHz. The meter integratesvalues of apparent conductivity over obser-vation depths of 0 co 75 cm and 0 co 150cm in the horizontal dipole and the verri-cal dipole orientations, respectively.

Study area

The study location was at the Mis-souri Management Systems EvaluationArea (MSEA) near Centralia, Missouri.The srudy site was in a fallow researchplot on a west-facing backslope of abroad, low, upland divide with relativelylow relief« 2.5 m).

The srudy site was located in an areaof Mexico (fine, montmorillonitic,mesic Udollic Ochraqualfs) soils (18)and included both eroded and overwashphases of Mexico soils on I % to 3%slopes. Mexico soils formed in moder-ately-fine texrured loess over fine tex-rured till. Depth co water table is verydeep (> 180 cm). The mineralogy isdominated by smectitic clays. Typically,the surface layer is silt loam, bur rangesfrom silt loam co silty clay loam depend-ing on the degree of erosion and subsoilmixing by cultivation. The subsoil issilty clay loam, silty clay, or clay. It isnot uncommon co have subsoil or clay-pan horizons with 50 co 60 percenr clay.Because of differences in clay content,contrasts in electrical conducrivity wereassumed to exist between the surface

and claypan horizon(s) of Mexico soils.Depth co claypan varied because of ero-

sion, deposition, and landscape position.Within [he srudy site, the depth to clay-pan ranged from less than 10 cm to morethan 100 cm. Variations in the thickness

of the surface and subsurface layers andthe contrasts in electrical conductivity be-tWeen these layers and rhe claypan were

Figure 2. Depth to claypan (A) and apparent conductivity (B) along study transect

considered rhe principal facrors likely coinfluence rhe EM response.

Field methods

A 190m transecr line was established

along the soUth border of plot 7 at theMissouri MSEA research site. Thirty-oneobservation poinrs were flagged at 6.3 mintervals along rhis rransect line. Ar eachobservarion point. the deprh co claypanwas determined with a soil probe (4.1 cmdiameter, 110 cm long). A characrerisricsharp increase in day content at the uppersurface of the daypan was recognized byits resistance COrhe hand-probe and tex-rural differences observed in the exrracred

cores. At each observation poine. rhe

EM38 meter was placed on the groundsurface and measurements were taken in

both horizontal and verrical dipole orien-tations. Depth to claypan measurementsand EM data collected at each observation

point along this line were compared andused ro develop regression equations ropredicr the depth ro claypan from valuesof apparent conductivity.

To demonstrate the use of EM rech-

niques and developed regression equationsco map the ropography and depict varia-tions in rhe depth ro the daypan, a small,12.2 by 18.3 m grid was esrablishedacross the ploe. The grid interval was 3.05m. Survey flags were inserted in theground ar each of rhe 35 grid intersec-

~()\'F\IBER-[)ECE\IBFR I"".. 573

20

00 20 40 60 80 100 120 140 160 180

Distance from East Edge of Plot (m)

70

- 60E-.en.§ 50

'S.+:; 40(,):J

"0

§ 30U...c:e20coQ.c.« 10

I

EM38Horintal DipoleEM38V$' Dipole0 I . I . , . I . I . , . , . I

0 20 40 60 80 100 120 140 160 180

Distance from East Edge of Plot (m)

rions. Wirh rhe EM38 merer placed onrhe ground surface, measuremenrs wereraken in both horizontal and vertical di-

pole orientarions at each grid intersection.Using rhe sofrware program SURFER,rwo- and rhree-dimensional simulations

of rhe site were constructed to help sum-marize predicred variations in the depthto claypan.

Results and discussion

Along the rransect line, the depth toclaypan averaged 39.6 cm and rangedfrom 7 cm to 105 cm (Figure 2A). Thedepth ro claypan was shallow «50 cm) ar71 % and moderarely deep (50 ro 100 cm)ar 26% of the observation points.

Values of apparent conducrivity aver-aged 32 mS/m (milliSiemens/merer) andranged from 18 mS/m ro 44 mS/m in rhehorizontal dipole orientation (Figure 2B).In rhe vertical dipole orientation, valuesof apparent conductivity averaged 54mS/m and ranged from 36 mS/m ro 68mS/m (Figure 2B). The higher readingsin the vertical dipole orientation impliedthat apparent conductivity increased withsoil depth. This relationship was believedto be a manifestation of increasing clayand moisrure contents with depth.

Similar patterns existed in both depthto the claypan and apparent conductivity(Figures 2A and 2B). Generally, depth toclaypan and apparent conductivity wereinversely related (Figure 3). The coeffi-

, , , ,\. ,. , ,

\' ' A

A', .......

\'}.,..,

..~. ,, ," ," ," ,.. ," ," , A

" ,"",,' Jr..~,."

A " ,AAA' , .

120

100

.E.2.80C~>-.!! 60(J0--= 40a.CDQ

.

.... .20

.. .,.... ... ..010 20 30 40 50 60

Apparent Conductivity (mS/m)80

Unear Regression

A Exponential Regression

Quadratic Regression

vertical Dipole

Horizontal Dipole.

70

Figure 3. Relationship between measured depth to claypan and apparent conduc-tivity values for study transect. Values of apparent conductivity collected with theEM 38 meter in the vertical dipole orientation. For regressions, refer to equationsequations [1], [2], and [3] in text

cient of determination, r', betWeen thedepth ro claypan and values of apparentconductivity was 0.635 (significant at the0.005 level) in the horizontal dipole ori-entation and 0.727 (significant at the0,005 level) in the vertical dipole orienta-tion. Considering the large volume of soilmeasured with the EM38 meter to the

A

<:> ""'"..

~.,.;» ""'..?

~~ ""'.:;>~ Ci>

.q t5'

~~ "..

~<S>0

~

~ -<:?:-~

~ ~ ~,,<t;;1>0 ~~

~ .c."""'~~ :§;:-.G ~

-<y"""'~"~

volume of soil sampled with the probe,the correlation was considered favorable.

At a given location, values of apparentconductiviry obtained with the differentcoil orientations were strongly interdepen-dent (r = 0.928). Since a stronger relation-ship existed betWeen the apparent conduc-tivity values obtained with the EM38

18

B

\ V /11//

('

/~(:ri\~!~ "o"-J , )

I( ~50/ ~6'0 ~\ '

~\60 ~

~~\ ,~

'" \.../ ~ \

>\\ / '\

\ \ o;,/~\j80. \ 1'; .?Rl~\1

16

U)0:::14~Eo-<~ 12:::a

Z 10-~UZ<t:Eo-<U}-Q

8

6

4

2

00 2 4 6 8 10 12

DISTANCE IN METERS

Figure 4. Three-dimensional surface net diagram of the claypan surface (A);Two-dimensional plot of the depth to the claypan (B);For each simulation, the contour interval is 10 cm

574 IOCR:\AL OF SOIL AND WATER CO:\SERVATION

.. ....

meter' in the vertical dipole orientation andthe depth co claypan, these values wereused co develop regression equations copredict depth co claypan from apparent

conducrivity. The following equationsweredeveloped for the srudy site:

Linear regression(r' =0.73):D= 194.9-2.878 X

Quadratic (r' = 0.77):D = 417.0 - 11.68X + 0.0839 X' [2]

Exponential (r' = 0.81):D = 1662 10 -''',j3X

where D is depth co claypan (cm) andX is the apparent conductiviry measuredby the EM38 meter in the verrical dipoleorientation (mS/m). An exponential re-gression equation appeared ro be the bestchoice co predict the depth co claypanfrom values of apparent conductiviry col-lected with the EM 38 meter in the verri-

cal dipole orientation (Figure 3).A large number of measurements can

be collected with an EM meter in a veryshorr period of time. Generally, at eachobservation point, EM measurementswere recorded in less than 30 seconds.

Empirical relationships becween the depthco claypan and EM response can be devel-oped from a limited number of observa-tion points. Regression equations can beused co predict the depth co claypan overlarge areas.

Large quantities of data are needed rosupporr soil information and modelingsystems. Some of this informaiion can be

quickly and efficienrly inferred with EMtechniques. Large amounts of EM datacan be processed and displayed on three-dimensional surface net diagrams (Figure4A) or cwo-dimensional plors (Figure 4B).

Depths ro claypan were estimated overa survey grid using EM38 verrical dipoledata and the exponemial regression equa-tion [3]. Figure 4A is a three-dimensionalsurface net diagram of the ropography ofthe claypan surface. The verrical scale hasbeen exaggerated aboUt 7.5 times. Figure4B is a cwo-dimensional plot of the depthto claypan within the grid site. In bothFigure 4A and 4B, the contour interval is10 cm. These computer simulations canbe used co summarize variations in the

depth ro the claypan, directions of lateralwater flow, and facilitate the location of

sampling sites. However, the accuracy ofthese simulations depends on the adequa-cy of the data used ro construct the regres-. .Slon equanons.

Conclusions

An EM technique was found co be use-ful for determining the depth co claypan

[1]

at the Missouri MSEA srudy site. A highcorrelation was found becween EM re-

sponse and the depth co claypan in anarea of Mexico soils under similar man-

agement practices. More investigations inother areas of Mexico soil are needed ro

establish the general form of the relation-ship. In order ro assess the effects of dif-fering soil moisrure and temperarure con-ditions, investigations need co includemore extensive data sees collected at mul-

tiple claypan-soil sites and at differenttimes of the year. It appears feasible, atleast in some areas, ro use EM methods

and computer graphics techniques formapping the depth to contrasting soilhorizons and lithologic layers, thus en-hancing our understanding of these sub-surface fearures.

[3]

REFERENCES CITED

1. Ammons. J.T.. M.E. Timpson. and D.L. New-ron. 1989. Application of an aboveground electro-magnetic conductivity meter to separate Na-traqualfi and Ochraqualfi in Gibson Counry,Tennessee. Soil Survey Horizons 30(3):66-70.

2. Cook, P.G., M.W. Hughes, G.R. Walker, andG.B. Allison. 1989. The calibration of frequenry-domain ekctromagnetic induction meters and theirpossibk me in rechargesrudies. Journal of Hydrol-ogy 107:251-265.

3. Cook, P.G., G.R. Walker, G. Buselli, I. Pores,

and A.R. Dodds. 1992. The application of ekc-tromagnetic techniques to groundwater rechargeinvestigations. Journal of Hydrology 130:201-229.

4. Corwin, D.L. and J.D. Rhoades. 1982. An im-proved technique for determining soil electricalconductivity-depth relations from above-groundekctromagnetic measurmzents. Soil Science Soci-ety of America Journal 46:517-520.

5. Corwin. D.L. and J.D. Rhoades. 1990. Estab-lishing soil electrical conductivity-depth relations

from electromagnetic induction measurements.Communications in Soil Science Plant Anal.21(11&12):861-901.

6. De Jong. E., A.K. Ballantyne. D.R. Cameron,andD.W. L. Read. 1979.A1easurementofappa~ent ekcrrical conductivity of soils by an electromag-netic induction probe to aid salinity SItrveys.SoilScience Society of America Journal 43:810-812.

7. Greenhouse. J.P.. and D.D. Siaine. 1983. The useof reconnaissance ekctromagnetic methods to mapcontaminant migration. Ground Water Moniror-ing Review 3(2):47-59.

8. Kachanoski. R.G.. E.G. Gregorich. and I.J. VanWesenbeeck. 1988. Estimating spatial variationsof soil water content using noncontacting ekctro-magnetic inductive methods. Canadian Journal ofSoil Science 68:715-722.

9. Kawasaki. K. and T.E. Osterkamp. 1988. A1ap-ping shallow permafrost by ekctromagnetic induc-tion-practical considn-ations. Cold Region Sci-ence and Technology 15: 279-288.

10. McBride, R.A., A.M. Gordon. and S.c. Shrive.

1990. Estimating forest soil quality from terrainmeasurements of apparent electrical conductivity.Soil Science Sociery of America Journal 54:290-293.

11. McNeill. J.D. 1980. Ekcrromagnetic terrain con-ductivity measurements at low induction numbers.Technical Note TN-6. Geonics Ltd.. Missis-

sauga. Omario. p. 15.12. McNeill. J.D. 1986. Geonics EA138 ground con-

ductivity meter operating insm/ctions and surveyinterpretation techniques. Technical Note TN-21. Geonics Ltd.. Mississauga. Omario. p. 16.

13. Palacky, G.J. 1987. Clay mapping using elecuo-magnetic methods. First Break 5(8):295-306.

14. Rhoades, J.D~ and D.L. Corwin. 1981. Deter-

mining soil electrical conductivity-depth relationsusing an inductive ekcrromagnetic soil conductivi-

ty m(ter. Soil Science Society of America Journal45:255-260.

15. Rhoades. J.D.. P.A. Raats, and R.J. Prather.1976. Efficts of liquid-phase electrical conductivi-ty. water content. and suiface conductivity on bulksoil electrical conductivity. Soil Science Sodety ofAmerica Journal 40:651-655.

16. SJavich, P.G. and G.H. Petterson. 1990. Esti-

mating average rootzone salinity from ekcrromag-netic induCtion (EA1-38) measuremenrs. Aus-tralian Journal of Soil Research 28:453-463.

17. Smith, D.D. and D.M. Whitt. 1947. Estimatingsoil lossesfrom field areas of claypan soiL Soil Sci-ence Society of America Proc. 12:485-490.

18. Soil Survey Staff. 1992. Keys to Soil Taxonomy.5th edition. SMSS technical monograph No. 19.Blacksburg. Virginia. Pocahontas Press. 1nc.. p.556. .

19. Williams, B.G. and G.c. Baker. 1982. An ekc-

tromagnetic induction technique for reconnaissancesurveys of soil salinity hazards. Australian Journalof Soil Reearch. 20:107-118.

20. Zalasiewicz. JA, S.J. Mathers, and.J.D. Corn-well. 1985. The application of ground conductivi-ty measurmzents to geological mapping. QuatterlyJournal of English Geology. London 18:139-148.

Single reprinrs of articles from the Journal ofSoil and Water Comervation are availablefrom the Soil and Water Conservation Soci-

ety, 7515 Northeast Ankeny Road, Ankeny,Iowa 50021-9764, at a cost of $0.25 perpage or a minimum of $5 per order.

,O\.F\IBFR.J)ICF\IBFR 1""4 575