5EvapotranspirationR.A.FeddeslandK.J.Lenselink25.1IntroductionEvapotranspirationisimportantasaterminthehydrologicalcycle,e.g.insoilwaterandgroundwaterbalances(Chapter16),andinsalinization(Chapter15).Inlanddrainageengineering,wethereforeneedtodevoteproperattentiontoitsdetermination,particularlyinaridandsemi-aridareas.Thisappliesnotonlytothevarioussurveysandinvestigationsthatprecedeadrainagedesign,butalsotothesubsequentmonitoringoftheeffectsofdrainagemeasuresonparameterslikewatertabledepth,soilsalinity,and,ultimately,oncropyield.Inaddition,agriculturistswanttohaveinformationontheeffectsofawatersupplyoncropproduction.Asthereisoftenadirectrelationbetweentheratioofactualtopotentialevapotranspirationandactualtopotentialcropyield,agriculturistswanttoknowthespecificwaterrequirementsofacrop,andwhethertheserequirementsarebeingmetundertheprevailingenvironmentalconditions.Regularestimatesofevapotranspirationmayrevealwatershortagesand/orwaterlogging,whichcanthenleadtotechnicalmeasurestoimproveirrigationanddrainage,and,againultimately,toanincreaseincropyields.Thischapter,afterexplainingsomebasicconcepts(inSection5.2),providesbriefinformationonhowtomeasureactualevapotranspirationinthefieldandonhowtoestimatetheevaporativedemandoftheatmosphere.Actualevapotranspirationcanbemeasuredwiththesoilwaterbalanceapproach,orwithmicro-meteorologicalmethods.ThesewillbebrieflydiscussedinSection5.3.Actualevapotranspirationcanalsobeestimatedwithcomputermodelsorremote-sensingtechniques(Section5.6.4).Afewempirical,temperature-basedmethodsforestimatingpotentialevapo-transpirationarebrieflydiscussed(Section5.4).ThetheoryofPenman'sopenwaterevaporationistreatedfairlyextensivelyinSection5.5.ThisisfollowedbytherecentlyacceptedPenman-Monteithmethodofestimatingthepotentialevapotranspirationfromcroppedsurfaces,distinguishingbetweenwetanddrycrops,betweenfullandpartialsoilcover,andbetweenfullandlimitedwatersupply(Section5.6).HowtheprecedingtheoryisappliedinpracticeisexplainedinSection5.7,withtheuseofareferenceevapotranspirationandcropcoefficients.5.2ConceptsandDevelopmentsInthepast,manyempiricalequationshavebeenderivedtocalculatepotentialevapo-transpiration(i.e.evapotranspirationfromcroppedsoilswithanoptimumwater'AgriculturalUniversity,WageningenInternationalInstituteforLandReclamationandImprovement145supply).Onlytwoofthesemethodswillbedescribed:onebasedonairtemperatureanddaylength(BlaneyandCriddle1950),andanotherbasedonairtemperatureandsolarradiation(Turc1954;JensenandHaise1963).Theseempiricalcorrelationmethodsareoftenonlyvalidforthelocalconditionsunderwhichtheywerederived;theyarehardlytransferabletootherareas.Nowadays,therefore,thefocusismainlyonphysically-basedapproaches,whichhaveawiderapplicability.Fortheprocessofevapotranspiration,threebasicphysicalrequirementsinthesoil-plant-atmospherecontinuummustbemet.Theremustbe:A:Acontinuoussupplyofwater;B:Energyavailabletochangeliquidwaterintovapour;C:Avapourgradienttomaintainafluxfromtheevaporatingsurfacetotheatmosphere.Thevariousmethodsofdeterminingevapotranspirationarebasedononeormoreoftheserequirements.Forexample,thesoilwaterbalanceapproachisbasedonA,theenergybalanceapproachonB,andthecombinationmethod(energybalanceplusheatandmasstransfer)onpartsofBandC.Penman(1948)wasthefirsttointroducethecombinationmethod.Heestimatedtheevaporationfromanopenwatersurface,andthenusedthatasareferenceevaporation.Multipliedbyacropfactor,thisprovidedanestimateofthepotentialevapotranspirationfromacroppedsurface.Thecombinationmethodrequiresmeasuredclimaticdataontemperature,humidity,solarradiation,andwindspeed.Becauseeventhiscombinationmethodcontainsanumberofempiricalrelationships,numerousmodificationstoadjustittolocalconditionshavebeenproposedbyahostofresearchers.Analyzingarangeoflysimeterdataworldwide,DoorenbosandPruitt(1977)proposedtheFAOModifiedPenmanmethod,whichhasfoundworldwideapplicationinirrigationanddrainageprojects.Theseauthorsadoptedthesametwo-stepapproachasPenmantoestimatecropwaterrequirements(i.e.estimatingareferenceevapo-transpiration,selectingcropcoefficientspercropandpergrowthstage,andthenmultiplyingthetwotofindthecropwaterrequirements).TheyreplacedPenman’sopenwaterevaporationbytheevapotranspirationfromareferencecrop.ThereferencecropofDoorenbosandPruittwasdefinedas‘anextendedsurfaceofan8to15cmtallgreengrasscoverofuniformheight,activelygrowing,completelyshadingtheground,andnotshortofwater’.Therewasevidence,however,thatthemethodsometimesover-predictedthecropwaterrequirements.UsingsimilarphysicsasPenmandid,Monteith(1965)derivedanequationthatdescribesthetranspirationfromadry,extensive,horizontally-uniformvegetatedsurface,whichisoptimallysuppliedwithwater.Ininternationalliterature,thisequationisknownasthePenman-Monteithequation.InTheNetherlands,thenameofRijtemahasbeenadded,becausethisauthorindependentlyderivedasimilarformula(Rijtema1965).Recentcomparativestudies(e.g.thosebyJensenetal.1990,whoanalyzedvariousmethodsofestimatingpotentialevapotranspiration)haveshowntheconvincingperformanceofthePenman-Monteithapproachunderv