Title 土壌マトリックポテンシャルセンサー MPS‑1を対象とした経験的な温度 校正式の作成
Author(s) 出口, 哲久; 岩間, 和人
Citation 北海道教育大学紀要. 自然科学編, 69(2): 9‑16
Issue Date 2019‑02
URL http://s‑ir.sap.hokkyodai.ac.jp/dspace/handle/123456789/10339
Rights
土壌マトリックポテンシャルセンサー MPS-1を対象とした 経験的な温度校正式の作成
出口 哲久・岩間 和人
＊北海道教育大学札幌校栽培・生物育成研究室
＊北海道大学・作物学研究室
DevelopmentofEmpiricalTemperatureCalibrationEquationfor SoilMatricPotentialSensorMPS-1
DEGUCHITetsuhisaandIWAMAKazuto
＊DepartmentofCultivationandNurturingLivingThings,SapporoCampus,HokkaidoUniversityofEducation
＊DepartmentofFieldCropScience,HokkaidoUniversity
ABSTRACT
Tomeasurethesoilmatricpotential(ψsoil)overawiderange,newtypesofsensorshave recently been developed. Among these sensors, MPS-1 seems especially suitable for replicatedfieldmeasurementbecauseofitssmallsizeandlowprice.However,MPS-1could possesstemperaturedependencybecauseofdielectricmeasurement.Theobjectiveofthis research was to evaluate the temperature dependency and to develop an empirical temperature calibration equation for it. The results demonstrate the existence of temperaturedependencyinMPS-1.Thetemperaturedependencycanbecomesevere, especiallyindrysoilconditions.Fromtheseresults,anempiricaltemperaturecalibration equationwasdeveloped.Usingtheequation,theeffectsoftemperaturedependencywere almostcompletelycalibrated.Thiscalibrationequationcanmakepossiblethecontinuous measurementofψsoilwithMPS-1infieldconditionswithoutanytemperaturedependency.
１．Introduction
Toevaluatethesoilwater–plantrelationship,measurementofsoilmatricpotential(ψsoil)isnecessary.
Becausemostplantscanabsorbwaterdawntoψsoilofapproximately−1.5MPa,itisdesirabletomeasure ψsoiluptothisrange.However,continuousandnondestructivemeasurementsofψsoilin situhavelong beenconductedwithawater-filledtensiometer,whoserangeofmeasurementislimitedto0to−85kPa.
出口 哲久・岩間 和人
Manyalternativemethodswithawidermeasuringrangehavebeenproposed,e.g.,electricalresistance sensorsorthermocouplepsychrometers,butmanyconstraintsstillexist,suchasaccuracy,hysteresis, andtechnicalrequirementsforusers[1].
Recently,newsensorsusingelectromagnetictechniqueshavebeenreported[2-5],andseveralsensors usingsimilarconceptsarecommerciallyavailabletoday(e.g.,MPS-1,DecagonDevices,Inc.,Pullman, WA,USA;EQ-2,Delta-TDevices,Cambridge,England;EQ-15,EcomatikGmbH,Dachau,Germany).
Thesesensorscanmeasureawiderangeofψsoil,andtheydonotneedmaintenanceforrefillingor degassingasawater-filledtensiometerdoes.
However,involumetricwatercontentsensorsbasedondielectricmeasurements,dependenceonsoil temperaturehasfrequentlybeenreported[6,7].Thisdependenceismainlyduetoanegativetemperature dependencyofthedielectricconstantofwater.Itisalsoaffectedbyelectricalconductivity[8]andsoil texture[9].
Therefore,ψsoilsensorsbasedondielectricmeasurementmayalsobeaffectedbysoiltemperature.
In this report, the temperature dependency of MPS-1 was examined under different soil water conditions.Then,anempiricaltemperaturecalibrationequationofMPS-1wasdeveloped.
２．Materials and Methods
２．１．Sensor’s characteristics
ThebodyofMPS-1consistsofanequilibriummediumthatequilibratesinψsoilwiththesurrounding soil.Inaddition,thewatercontentinsidetheequilibriummediumcanbemeasuredbasedonelectrical conductivity.Thewaterretentionpropertiesintheequilibriummedium(Fig.1)arerepresentedasthe followingequation.
ψsoil＝−exp(6.43×10−6×raw2−3.10×10−2×raw＋39.45) (1) where“raw”meansthesensor’soutputinrawcount.Thewaterretentionpropertywasdeterminedina pressureplateat22℃.Thesensor’soutputwasrecordedbyProcheck(DecagonDevices).
２．２．Experimental condition
Toevaluatetemperaturedependencyofsensor’soutput,measurementinvarioustemperatureunder stablesoilwatercontentisrequired.
Inthisexperiment,MPS-1wasinstalledinasoil-packedPVCpipewithaninnerdiameterof5.1cm andaheightof10cm(Fig.2).SoilintheplowlayerinanexperimentalfieldatHokkaidoUniversitywas airdriedandsieved(＜2mm).Then,differentamountsofwaterwereaddedtomakedifferentlevelsof θg(15,20,25,30,35,and40%).IneachPVCpipe,soilswerepackedatabulkdensityof1.0gcm-3.Two PVCpipeswereusedforeachθg.Beforetheinstallation,MPS-1wasdippedintopurewaterforoneday.
Toavoidwaterlossduringmeasurement,thebottomofthePVCpipewaswrappedtwicewithparafilm (PARAFILMM;PechineyPlasticPackaging,Chicago,IL,USA).BecausethecodeofMPS-1inhibited wrappingonthetopofPVCpipe,siliconerubber(siliconesealantCemedine8090pro;CemedineCo.,Ltd., Tokyo,Japan)wasused.Inaddition,tomeasuresoiltemperature,otherPVCpipeswiththermistor
sensorswereprepared.ThesoilwatercontentsofthePVCpipeswere10and40%.Outputsofthermistor sensorswererecordedwithanelectronicthermometer(SK-L200T,SatoKeiryokiMfg,Tokyo,Japan).
Beforethestartofmeasurement,equilibrationofsoilsandtheequilibriummediumofMPS-1were necessary.Therefore,thePVCpipeswereplacedinatemperaturechamberat42℃.However,inthe silicone layer, some clearances were observed because of the shrinkage of silicone. To fill these clearances,siliconewasaddedagain.Then,thesensor’soutputswerecheckedonceadayandconfirmed withequilibration.
After the outputs became stable, the outputs and soil temperature were recorded at 4℃ in a refrigerator,18 ℃and30 ℃inroomswithanairconditioner,and42 ℃inatemperaturechamber.To ensure that each PVC reached the same temperature, the outputs were recorded only when the differenceofsoiltemperaturebetween10and40%soilwatercontentswaslessthan1℃.Then,to measuretheactualθgofeachPVCpipe,thebottomwrapwasremoved,andsoilsinsidethepipewere manuallycollected(＜100g).
Inaddition,tocheckthewaterlossduringtheexperiment,asimilarPVCpipeof45%θgwasprepared withtworeplications.Then,thesameprocedureastheaboveexperimentwasrepeated,andtheweights ofthePVCpipesweremeasuredatthebeginningandtheendofmeasurement.Becausetheaverage weightlosswasonly0.07g(0.04%asθg),thewaterlossduringmeasurementwasconfirmedtobenegligible.
３．Results and Discussion
３．１．Temperature dependency of MPS-1 for different soil moisture contents
Ateachsoilwatercontent,thesensor’soutputwaslinearlyaffectedbytemperature(Fig.3).This resultindicatestheexistenceofatemperaturedependencyinMPS-1.Inaddition,thetemperature dependencydifferedamongθg.
Whenconvertedtoψsoil,theerrorresultingfromtemperaturedependencywassmalleratahighersoil Figure 1. Water retention property of MPS-1.
Equationofthewaterretentionpropertyis presentedbyDecagonDevices.
]=−exp(6.43×10−6×raw2−3.10×10−2×raw
＋39.45) Figure 2. DiagramofPVCpipeandMPS-1.
出口 哲久・岩間 和人
water content, while larger at a lower soil water content (Fig. 4). This result was due to the characteristicsofthewaterretentionpropertyofMPS-1.Atlowψsoil,asmallchangeofthesensor’s
Figure 3. Effectsoftemperatureonsensor’soutputindifferentsoil watercontent.
Eachsymbolindicatesgravimetricsoilwatercontent(%).
Verticalbarindicatesstandarderror(n=2).
Figure 4. Effectsoftemperatureonmatric potentialmeasuredwithMPS-1in different gravimetric soil water content.
Eachsymbolindicatesgravimetric soilwatercontent(%).
Vertical bar indicates standard error(n=2).
outputeasilyresultedinalargechangeofψsoil(Fig.1).
Ontheotherhand,apositivetemperaturedependencyseemedtoconflictwithanegativetemperature dependencyofwatermolecules.However,temperatureeffectsonsoildielectricpropertiesrelatenotonly towatercontentbutalsotosoiltextureandionbalance[10,11].Onceawatermoleculeisabsorbedtothe surfaceofsoilparticle,thewatermoleculebecomes“invisible”todielectricmeasurement.Whenthesoil temperatureincreases,theabsorbedwatermoleculeisreleasedfromthesoilparticleandbecomes
“visible”todielectricmeasurement.Especiallyinfine-texturedsoils,thenumberofabsorbedwater moleculesbecomeslargerbecauseofthehighsurfacearea.Therefore,inafine-texturedsoil,theeffectof thereleasedwatermoleculesonthedielectricconstantoftenexceedsthedepressionofthedielectric constant of the water molecules. Therefore, a positive temperature dependency of soil dielectric properties is a common feature in fine-textured soil. Because the measuring object of MPS-1 is equilibriummedium,whichismadebyceramicmaterialwithawideporesizedistribution,thepositive temperaturedependencyindicatesahighsurfaceareaofthemedium.
３．２．Development of empirical temperature calibration equation for MPS-1
Basedontheresults,thesensor’soutputwasdefinedasraw(S,T),becauseitdependedonthedegree ofsaturationinsidetheequilibriummedium(S)andtemperature(T).Inaddition,thetemperature dependencywasdefinedasa(S)becauseitdependedonS.
Tocalibratetheeffectoftemperature,raw(S,T)needstobecalibratedtothesensor’soutputat22℃, thetemperaturewhenthewaterretentionpropertyisobtained.Tocalibrateraw(S,T)toraw(S,22), followingequationisused:
raw(S,22)＝raw(S,T)−(T−22)×a(S) (2)
Becausetheeffectivemeasurementrangeof MPS-1was −10to −500kPa,Swasdefinedas1at
−10kPaand0at−500kPa.Fromagivensensor’soutput,whichcanbeexpressedasraw(S,T),Scan becalculatedbythefollowingequation:
S＝ raw(S,T)−raw(0,T)raw(1,T)−raw(0,T) (3)
FromEquation(2),raw(1,T)andraw(0,T)wereobtainedbythefollowingequation:
raw(1,T)＝raw(1,22)＋(T−22)×a(1)
(4) raw(0,T)＝raw(0,22)＋(T−22)×a(0)
Inaddition,becauseEquation(1)wasobtainedat22℃,raw(1,22)andraw(0,22)couldbeobtainedby calculatingtherawat−10and−500kPa,respectively.Asaresult,thefollowingequationwasobtained:
raw＝(1,22)＝2227.08
(5) raw＝(0,22)＝1609.28
FromthelinearexpressionobtainedfromFig.3,a(S),raw(S,0),raw(S,22),andSatdifferentsoil watercontentsaresummarizedinTable1.Generally,soilwatercontentswerelowerthanintended
出口 哲久・岩間 和人
values.ThiswasduetoevaporativelossfromtheclearancebetweenthesiliconeandMPS-1.Evaporative lossmightalsooccuratsoilpacking andcollecting times. However, because thewaterlostduring measurementwasnegligible,θginPVCwasassumedtobestableduringevaluationoftemperature dependency.FromthedataofTable1,therelationshipbetweenSanda(S)wasobtainedasthefollowing equation(Fig.5).
a(S)＝2.5474S＋1.262 (6)
FromEquation(6),a(1)anda(0)werecalculatedasbelow:
a(1)＝3.8094
(7) a(0)＝1.262
Bysubstituting Equations(4),(5),and(7)intoEquation(3),thedegreeofsaturationcouldbe calculatedonlybythesensor’soutputandtemperatureusingthefollowingequation:
S＝raw(S,T)−1581.516−1.262T561.7572＋2.5474T (8)
Subsequently,bycombiningEquations(2),(6),and(8),acalibratedsensor’soutputwasobtainedbya singleequation:
raw(S,22)＝raw(S,T)＋(22−T)
(
2.5474×raw(S,T)−1581.516−1.262T561.7572＋2.5474T ＋1.262）
(9)Finally,bysubstitutingthecalibratedsensor’soutputintoEquation(1),calibratedψsoilwasobtained.
Usingthiscalibrationequation,thedataofFigs.3and4werecalibrated(Figs.6and7).Asaresult, regardlessofθg,theeffectofthetemperaturedependencywasalmostcompletelycalibratedwiththis calibrationequation.
Figure 5. Relationship between degree of saturationandthermaldependency.
＊＊indicatessignificanceatP＜0.01.
Table 1 . Thermal dependency a(S), sensor’s output at 0℃ and 22℃ (raw (S,0), raw (S,22)) degree of saturation (S) at differentgravimetricsoilwat
４．Conclusions
Inthisresearch,thetemperaturedependencyofMPS-1wasconfirmed.Theerrorcausedbythe temperaturedependencybecomesseverewhenthedegreeofsaturationintheequilibriummedium becomeslower.
Fromthedatacollectedinthiswork,anempiricalcalibrationequationwasdeveloped.Withthis equation,thetemperature-derivederrorwassuccessfullycalibrated.
TousesensorsbasedonamechanismsimilartoMPS-1,researchersneedtoconsidertemperature dependency.Also,thedevelopersofsensorsshouldnoteinaninstructionmanualthetemperatureat whichthewaterretentioncurveoftheequilibriummediumisobtained.
５．Acknowledgments
TheauthorswouldliketothankJ.Kashiwagiforusefulsuggestionsforexperimentaldesign.
６．References
［1］PardossiA,IncrocciL,IncrocciG.etal.2009.Rootzonesensorsforirrigationmanagementinintensiveagriculture.
Sensors9:2809-2835.
［2］HilhorstMA,DeJongJJ.1988.Adielectrictensiometer.AgriculturalWaterManagement13:411-415.
［3］WhalleyWR,WattsWR,HilhorstMAetal.2001.Thedesignofporousmaterialsensorstomeasurethematric Figure 6. Effect of temperature calibration on
sensor’soutputshowninFigure3.
Each symbol indicates gravimetric soil watercontent(%).
Vertical bar indicates standard error (n=2).
Figure 7. Effect of temperature calibration on matricpotentialshowninFigure4.
Each symbol indicates gravimetric soil watercontent(%).
Vertical bar indicates standard error (n=2).
出口 哲久・岩間 和人 potentialofwaterinsoil.EuropeanJournalofSoilScience52:511-519.
［4］WhalleyWR,LochG,JenkinsM,etal.2009.Measurementoflowmatricpotentialswithporousmatrixsensorsand water-filledtensiometers.SoilScienceSocietyofAmericaJournal73:1796-1803.
［5］MurrayL.BingC.S.,2008.CoiledTimeDomainReflectometryMatricPotentialSensor.SoilScienceSocietyofAmerica Journal72:1422-1424.
［6］BaumhardtRL,LascanoRJ,EvettSR.2000.Soilmaterial,temperature,andsalinityeffectsoncalibrationofmultisensor capacitanceprobes.SoilScienceSocietyofAmericaJournal60:1940-1946.
［7］KizitoF.,CampbellC.S.,CampbellG.S.etal.2008.Frequency,electricalconductivityandtemperatureanalysisofa low-costcapacitancesoilmoisturesensor.JournalofHydrology352,367-378.
［8］PerssonM,Berndtsson.1998.Textureandelectricalconductivityeffectsontemperaturedependencyintimedomain reflectometry.SoilScienceSocietyofAmericaJournal62:887-893.
［9］Or D., Wraith J.M. 1999. Temperature effects on soil bulk dielectric permittivity measured by time domain reflectometry:Experimentalevidenceandhypothesisdevelopment.Waterresourcesresearch35:361-369.
［10］SayfriedMSMudockMD.2001.Responseofanewsoilwatersensortovariablesoil,watercontent,andtemperature.
SoilScienceSocietyofAmericaJournal65:28-34.
［11］SayfriedMSMudockMD.2004.Measurementofsoilwatercontentwitha50-MHzsoildielectricsensor.SoilScience SocietyofAmericaJournal68:394-403.
（出口 哲久 札幌校講師）
（岩間 和人 北海道大学名誉教授）
