智慧农业（中英文）SmartAgricultureVol.3,No.2
44
sponsecharacteristicsandchlorophyllcontentestima‐2013,33(3):766-769.
tionofphyllostachysviolascensleavesunderdroughtFUY,FANY,SHENGJ,etal.Studyonrelationship
stress[J].ActaEcologicaSinica,2018,38(18):322-329.betweenalfalfacanopyspectralreflectanceandleaf
[37]付彦博,范燕敏,盛建东,等.紫花苜蓿冠层反射光谱watercontent[J].SpectroscopyandSpectralAnalysis,
与叶片含水率关系研究[J].光谱学与光谱分析,2013,33(3):766-769.
IdentificationandLevelDiscriminationofWaterlogging
StressinWinterWheatUsingHyperspectralRemoteSensing
YANGFeifei,LIUShengping,ZHUYeping,LIShijuan
（AgriculturalInformationInstitute,ChineseAcademyofAgriculturalSciences/KeyLaboratoryofAgri-information
ServiceTechnology,MinistryofAgricultureandRuralAffairs,Beijing100081,China）
Abstract:Thefrequentoccurrenceofwaterloggingstressinwinterwheatnotonlyseriouslyaffectsregionalfoodsecurityand
ecologicalsecurity,butalsothreatenssocialandeconomicstabilityandsustainabledevelopment.Inordertoidentifythewater‐
loggingstresslevelofwinterwheat,awaterloggingstressgradientpotexperimentwassetupinthisresearch.Threefactors
werecontrolled:waterloggingstresslevel(control,slightwaterlogging,severewaterlogging),stressduration(5days,10days,
15days)andwheatvariety(YF4,JM31,JM38).LeafandcanopyhyperspectraldataweremeasuredbyusingASDFieldSpec3
andGaiasky-mini2imagingspectrometer,respectively.Thedatawerecollectedfromthefirstwaterloggingdayofwinterwheat.
Thesunnyandwindlessweatherwasselectedandmeasuredevery7daysuntilthewheatwasmature.Combinedwithvegeta‐
tionindex,normalizedmeandistanceandspectralderivativedifferenceentropy,ifwinterwheatwasunderwaterloggingstress
wasmonitoredandstresslevelwasidentified.Theresultsshowedthat:1)thespectralresponsecharacteristicsofwinterwheat
underwaterloggingstresschangedsignificantlyinRW,RE,NIRand1650－1800nmregion,whichmaybeduetothesensitivi‐
tyoftheseregionstophysiologicalparametersaffectingthespectralresponsecharacteristics,suchaspigment,nutrient,leafin‐
ternalstructure,etc;2)thesimpleratiopigmentindexSRPIwastheoptimalvegetationindexforidentifyingthewaterlogging
stressofwinterwheat.Theexcellentperformanceofthisvegetationindexmaycomefromitsextremesensitivitytotheepoxida‐
tionstateandphotosyntheticefficiencyofthexanthophyllcyclepigment;3)theredlightabsorptionvalley(RW:640－680nm)
regionwastheoptimalregionforidentifyingwaterloggingstresslevel.InRWregion,waterloggingstresslevelofwinterwheat
couldbedeterminedbythespectralderivativedifferenceentropyatheading,floweringandfillingstages.Thegreaterthelevel
ofwaterloggingstress,thegreaterthespectralderivativedifferenceentropy.ThismaybeduetothefactthattheRWregionwas
moresensitivetopigmentcontent,andthespectralderivativedifferenceentropycouldreducetheeffectsofspectralnoiseand
background.Thisstudycouldprovideanewmethodformonitoringwaterloggingstress,andwouldhaveagoodapplication
prospectintheprecisepreventionandcontrolofwaterloggingstress.Therearestillshortcomingsinthisstudy,suchasthedif‐
ferencebetweenthepotexperimentandtheactualfieldenvironment,thelackofindependentexperimentalverification,etc.
Nextresearchcouldaddpotandfieldexperiments,combinewithcross-validation,tofurtherverifythefeasibilityofthisre‐
searchmethod.
Keywords:hyperspectralremotesensing;waterloggingstress;vegetationindex;spectralderivativedifferenceentropy;winter
wheat
（登陆www.smartag.net.cn免费获取电子版全文） |
|
