DOC

Application of principal component analysis to evaluation of black soil degradation in Jilin

By Gail Tucker,2014-11-19 05:15
8 views 0
Application of principal component analysis to evaluation of black soil degradation in Jilin

    Application of principal component analysis to evaluation of black soil degradation in

    Jilin

    GlobalGeology,14(1):5458(2011)

    doi:10.3969/j.issn.16739736.2011.01.07

    ArticleID:16739736(201I)01005405

    ---'-一??-一一一

    Applicationoiprincipalcomponentanalysistoevaluation

    ofblacksoildegradationinJilin

    NIUCencen,WANGQing,WENXiuyu,GU0Yuan,

    ZHAGNPenglei,ZHURongyueandHEXianhui

    CollegeofConstructionEngineering,JilinUniversity,Changchun130026,China Abstract:PrincipalComponentAnalysis(PCA)cansimplifythestructureofdatabasebyreplacingmulti.

    dimensionalparameterswithrelativelylesscomprehensivevariablesinordertoensuretheminimumlostinini

    timdata.Inthispaper,eighteenblacksoilsamplesfromdifferentsitesweretestedandthirteendistinctivein

    dexeswerechosentoevaluatethedegenerationofblacksoil.Byusingprincipalcomponentanalysis,variables

    ofthirteendimensionscanbediminishedtosixunrelatedprincipalindexes.AnalysisshowsthatthesolublesaIt

    content,Fulvicacids(FA)andaggregationdegreehaveahighweighingcoefficient.indicatingthesethreein

    dexesarethemajorpartsf0rtheevaluationofblacksoildegradation.Italsoprovidesanewpathtothedegener

    atedblacksoiltreatmentinNortheastChina. Keywords:principalcomponentanalysis;blacksoil;soildegradation

    1Intr0ducti0n

    Blacksoilisakindofsemisolublesoilcharac.

    terizedbyhighfertility,thickhumusandloosetex

    ture.BlacksoilsintheNortheastChina,whereisone ofthethreemajordistributionareasofblacksoilin theworldandanimportantgrainproductionbase,had beenindegradationduetotheincreasingofpopula

    tion.irrationalreclamationandsomeinevitablenatu

    ralreasons.Asaresult,thedegradationofblacksoil inNortheastChinahasbeguntorestrictthesocial andeconomicdevelopmentbylargelythreateningthe nation'sfoodsecurity.Therefore,itisnecessaryto keepresearchonthisissue(Xu,2005).

    Soildegradationwasfirstlyproposedasaspecial subjectbytheFoodandAgricultureOrganizationof theUnitedNations,andthencameintopeople'seye

    sightincludingChineseresearchers.WangJingkuan analyzedthecurrentsituationandchanginglawofor- ganicmatters,Nitrogen,SulfurandPhosphorusin blacksoilsfromvarioussites(Wang,2002).LYi

    zhongdiscussedtheeffectsofdifferenttillagemeas. uresonhumilicacidandtheinfraredspectrumfea

    tures(L0,2009).SUIYaoyutestedtherelationship oforganicmatters,andactivitiesinblacksoil(Suiet a1.,2009).ZOUWenxiusearchedtheregularityof watercontentchangingindifferenttypeoflandutiliza

    tion(Zoueta1.,2009).QINYanevaluatedthe structureofblacksoilinJilinProvinceusinggray

fuzzyclusteringmethod(Qineta1.,2008).Tosum

    mary,researchesonblacksoilwereprimarilycen

    teredonthechemicalcomponentsaswellasthecoin

    ponentsehangingwiththeenvironment.

    Inthispaper,toinvestigatethemechanismof Received12January2011,accepted28January2011 SupportedbyCollegeStudentsInnovationExperinmntPlanofJilinUniversity(No.2009A6

    3066);theNationalNaturalScienceFoundati0nofChina(N0 406721800);theInternationalCooperationProjeetsofNationalNaturalSeieneeFunds(No.

    40911120044)

    ApplicationofprincipalcomponentanalysistoevaluationofblacksoildegradationinJilin55

    blacksoildegradationinJilinProvince,principal componentanalysisisusedtosimplifythemassstruc

    tureindexesofblacksoilanddiminishthenumberof variables.

    2Soilsamplesandlaboratorystudies

    Referringthegeologicalmapoftheblacksoil distributioninJilinProvince(Fig.1),sixsitesindif- ferentformationandlithologywereselected:E125. 4339,N43.7630Jingyuetan;E125.7118,N43.5187 Shuangyang;N44.114,E125.407Mishazi;N43. 8815,E124.7564Huande:N44.4517,E126.1254 Dagou:N44.6540.E126.4357Xiushui.Threesam

    piesindifferentdepth(010cm,1020cmand20

    30cm)ofeachsite,totallyeighteensoilsamples, weretaken.

    Laboratorytestsinclude:thebasicphysical chemistrytests,whichcanget:contentofsolublesalt (A1),abilityofcationic(A2),organicmatters(A3), PHvalue(A4),HA(A5),FA(A6),humilicacid

    (A7);thegrainsizeanalysisandaggregationanalysis experiments,whichcanget:therateofstructural damage(A8),degreeofdispersion(A9),aggregation degree(A10);andthemercurypenetrationtestgives themedianaperture(AI1),percentofvoid(A12), fractaldimensionofsoilskeleton(Al3). Fig.1Distributionoftheblacksoilinthestudiedarea 3Derivationofprincipalcomponent

    Principalcomponentanalysisisanmthematics procedurethattransformsanumberofpossiblyeorre

    latedvariablesintoasmallernumberofuncorrelated parameterscalledprincipalcomponents(He,1998; Zhangeta1.,2010;Luo,2010;Wan,2007;Yaoet a1.,2010).Itsmajorproceduresareasfollowings: (1)Inputvariablesanddata.

    Suppose:thereare,vsoilsamplesandPindexes foreachsample.ThentheseN:l:Pdatacanbesigned inmatrixmarkedX:

    X=

    1l

    :

    ?

    1

    1

    ?

    :

    Ynp

    Inwhich:(i=1,2,3n,=1,2,3,P)isthe

    rawvalueoftheNo.JindexoftheNo.isample. (2)Standardizationofthedata.

56NiuC.C.,WangQ.,WenX.Y.,etal

    Toavoidtheeffectofdifierencesindimensions andmagnitudesofthedataduringcalculation,theav

    eragesandstandarddeviationsofmatrixXwerecaleu

    latedfollowedbythestandardizedmatrixofdata. ,7L--

    /~i=

    I

    1n

    x

    

    j)z](=1,2,,P)

    Y=(x~)/si(i=1,2,,n;=1,2,,p)

    Where:

    J

    isthemeanofj-linedata;Sjisthestandard deviationofdataoflinej;Y=(Y)forthestand

    ardmatrix.

    (3)Calculatethecorrelationcoefficientmatrix R.

    =

    r/~1i=1,

    (,=1,2,,p)

    (4)CalculatetheeigenvalueofthematrixR. SolvetheeigenvalueequationlRA,l=0,

    andworkouttheeigenvalueA(i:1,2,3..P),

    whileAl?A2?……..?A.?0,thecorresponding eigenvectorisLl,L2,.,J.

    (5)Calculatetheprincipalcomponentloading matrix.

    Accordingtostipulation:contributionrateto ,theteamadoptthefirst

    principalcomponents,thenconsistedtheprincipal componentsmatrix(F).f×,whilecolunlnvectorF

    (u=l,2,,)isthevprincipalcomponentsofthe

    blacksoilsamples.

    p

    =

    ?Yti,.(i=1,2,,n;j=1,2,,).i=i

    (6)Calculatetheweightofparameters. Figureoutthevarianceandweightofthecom

    nlonfactoraccordingtothematrix.

    =

    2

    a2s

    F):

    1,2,)

    Where:isthevarianceofthecommonfactor intheparameter;

    4

    B1istheweightoftheparamete

    Applicationexamples

    TaketheblacksoilinJilinprovinceasthesub

    jecttoresearchthemechanismofsoildegradation. Firstly,thethirteenindexesoftheeighteenblacksoil sampleswasstandardizedandthestandardizedmatrix wasgotten:()1813(i:1,2,18,J=1,2,

    13).Then,getthecorrelationcoefficientmatrixR: (r)8l3bymathematicalcalculationofthestandard

    izeddata()l8x13.

    Bysolvingtheeigenvalueequation,theeigenval? ue,contributionrateandaccumulatedcontribution ratearegottenshowninTable1.Fromthedatainthe table,wecanfindthattheaccunmlatedcontribution rateofAlA6is88.9%thatislargerthan85%

    whichisthegeneralcriteria.Therefore,thesesix principalcomponentsarenamedasthefirstprincipal component,thesecondprincipalcomponent,andso on.Thesenewparametersbrieflyrepresenttheorigi

    naldata,andtheyareindependenttoeachother. 5Summariesandconclusions

    Thispaperbuildsasimplemoldtoevaluatethe blacksoildegradationbyusingtheprincipalcompo

    nentanalysis.Sixprincipalcomponentsretaining 88.9%informationarecollectedbydimensionalityre

    ductionfromtheoriginaldata,makingtheanalysis clear,simpleandpersuasive.Besides,theprincipal componentanalysisalsocalculatedtheweighingc0ef_ ficientsofcontentofsolublesalt.FAandaggregation degree,whichinfluencedthedegradationofblacksoil inJilinprovinceprimarily.Ithasdrawnoutthefol

    lowingconclusions:

    (1)ItisobviousthatA11hasahighvalueinthe principalcomponentanalysis.ForF1,F4,F6,the loadvalueofAl1areashighas0.3446,0.6145and 0.4957respectively.Forthatreason,themedianap

    erture(Al1)playedanimportantroleinevaluating thesoildegradation.

    (2)Fromthedataofthesecondprincipalcorn

    ponent,itisfoundthatHAismoreinfluentia1.Soit

%

    <J

    8

    ^

    ?

    /

    A

    ?

    O

    a

    Lq

    e

    =

    ?

    ?

    }=

    ApplicationofprincipalcomponentanalysistoevaluationofblacksoildegradalioninJilin57

    Table1Correlationcoefficient(CC)

    Table2Eigenvalue,contributionrateandaccumulatiedcontributionrate

    Table3Loadingofeachindexandweighingcoefficientonthesixprincipalcomponents

    Note:?Frefet()thevariancesoftheConllllonfactor;WCreferstoweighingcoefficients

    58NiuC.

    C.,WangQ.,WenX.Y.,etal

    canbeconcludedthatthelosingofhumuscontributes totheblacksoildegradation.

    (3)Aggregationdegreeisthemainpartofthe thirdprincipalcomponent,andthepercentofvoid greatlyaffectedthefifthprincipalcomponent,showing thatthesoilstructurerepresentedthesoildegenera tion.

(4)FromtheTable3,itcanalsofindoutthat

    thecontributionrateofdifferentindexes.Byusingthe principalcomponentanalysis,theweightcoefficients of4l,A6andA10arecalculatedas0.0818,0.0812 and0.0808.Apparently,contentofsolublesalt, FA

    andaggregationdegreehaveahighweighingcoeffi cient,whichplayanimportantroleinsoildegradation inJilinProvince,whileotherindexesarelesseffec. tive

    Rererences

    HeXQ.998.Methodsandapplicationofmodernstatistical analysis.Beijing:People'sUniversityofChinaPress, 281305.(inChinese)

    LaYZ,CongWW,LianXJ.2009.Effectofdifferenttillage managementsonblackhumicfractionsandIRspectral feature.SpectroscopyandSpectralAnalysis,29:2642

    2645.(inChinesewithEnglishabstract)

    LuoXW,YaoHL.2010Amodeforcomprehensivelveva1

    uatingrockqualitybaseonprincipalcomponentanalysis anditsapplication.RockandSoilMechanics,31(Sup

    p1.):452455.(inChinesewithEnglishabstract) QinY,WangQ,ChenHE,eta1.2008.Syntheticalevalua

    tionofJilinProvinceblacksoi1structureonthebasisof grayfuzzyclusteringmethod.JournalofJilinUniversity: EarthScience晟觑ion,3S:841846.(inChinesewith

    Englishabstract)

    SuiYY,JiaoXG,GaoCS,eta1.2009.Therelationshipa

    mongorganicmattercontentandsoilmicrobialbiomass

andsoilenzymeactivities.ChineseJournalofSoilSci

    ence,4O:10361039.(inChinesewithEnglishabstract) WanXY,BaoWM,JingYD.2007.Studyonfloodsimilari

    tybasedonprincipalcomponentsanalysis.WaterRe

    sourcesandPower,25:3639.(inChinesewithEnglish

    abstract)

    WangJK,WangTY.GuanLZ.2002.Anapproachtothe changesofblacksoilquality(I)changesoftheindi

    cesofblacksoilwiththeyear(s)ofreclamation.Journal ofShenyangAgriculturalUniversity.33:43_47.(inChi

    nesewithEnglishabstract)

    xuXB.WangQ.2005.Thecurrentstatusandprospectsof researchinblackearthdegradationinnortheastChina. EarthandScience,33(Supp1.):588591.(inChinese

    withEnglishabstract)

    YaoYP,LixB,GongFQ,eta1.2010.Applicationof weightedMahalanobisdistancediscriminantanalysismeth

    odtoclassificationofrockmassquality.ChineseJournal ofRockMechanicsandEngineering,29(Supp1.):4119

    4123.(inChinesewithEnglishabstract)

    ZhangW,ChenJP,QinSw,eta1.2010.Applicationof FCMbasedonprincipalcomponentsanalysisindebris flowclassification.JournalofJilinUniversity:EarthSci

    enceEdition,40:368372.(inChinesewithEnglishab

    stract)

    ZouWX,HanXZ,LiLH,eta1.2009.Studyonsoilnois

    turecharacteristicofdifferentlandusetypesinblacksoil. JournalofSoilandWaterConservation,23:236239.(in

    ChinesewithEnglishabstract)

Report this document

For any questions or suggestions please email
cust-service@docsford.com