DOC

Spatial and temporal variability of daily precipitation in Haihe River basin, 1958-2007

By Paula Austin,2014-02-18 22:21
6 views 0
Spatial and temporal variability of daily precipitation in Haihe River basin, 1958-2007of,in,and,daily,Haihe,River,basin,river,Basin

    Spatial and temporal variability of daily

    precipitation in Haihe River basin,

    1958-2007

    J.Gcogr.Sci.2010,20(2):248260

    D0l:l0.10I'7,s1l442-010.0248.0

    Spatialandtemporalvariabilityofdaily

    precipitationinHaiheRiverbasin,1958-2007

    CHUJianting',,

    XIAJun',XUChongyu,LILuI-,WANGZhonggen

    1.KeyLaboratoryofWaterCycle&RelatedLandSurfaceProcesses.InstituteofGeographicSciencesand

    NaturalResourcesResearch,CAS,Beijing100101,China;

    2.GraduateUniversityofChineseAcademyofSciences,Beijing100039,China; 3.InstituteofPhysics&Meteorology,UniversityofHohenheim,Garbenstr.30,D-70955Stuttgart,Germany;

    4.DepartmentofGeosciences,UniversityofOslo,POBOX1047Blindern,N-0316Oslo,Norway

    Abstract:TheseasonaIvariabilityandspatiaIdistributionofprecipitationarethemaincause offloodanddroughtevents.ThestudyofspatialdistributionandtemporaItrendofprecipita- tioninriverbasinshasbeenpaidmoreandmoreattention.However,inChina,theprecipita

    tiondataaremeasuredbyweatherstations(WS)ofChinaMete0rologicalAdministrationand hydrologicaIraingauges(RG)ofnationaIandIocalhydrologybureau.TheWSdatausually haveIongrecordwithfewerstations,whiletheRGdatausuallyhaveshortrecordwithmore stations.TheconsistencyandcorrelationofthesetwodatasetshavenotbeenwelIunder- stood.Inthispaper.theprecipitationdatafrOm30weatherstationsfor1958-2007and248 raingaugesfor199-2004intheHaiheRiverbasinareexaminedandcomparedusingIinear

    regression.5-yearmovingaverage,Mann-Kendalltrendanalysis,Kolmogorov-Smirnovtest,

    ZtestandFtestmethods.TheresultsshowthattheannuaIprecipitationfrOmbothWSand RGrecordsarenormallydistributedwithminordifferenceinthemeanvalueandvariance.It isstatisticallyfeasibletoextendtheprecipitationofRGbyWSdatasets.Usingtheextended precipitationdata.thedetailedspatialdistributionoftheannuaIandseasonaIprecipitation amountsaswelIastheirtemporaltrendsarecalculatedandmapped.Thevariousdistribution mapsproducedinthestudyshowthatforthewholebasintheprecipitationof1958-2007has beendecreasingexceptforspringseason.Thedeclinetrendissignificantinsummer,and thistrendisstrongerafterthe1980s.TheannuaIandseasonaIprecipitationamountsand changingtrendsaredifferentindifferentregionsandseasons.Theprecipitationisdecreasing fr0msouthtonorth.frOmcoastaIzonetoinlandarea.

    Keywords:climatechange;spatialandtemporalvariabilityofprecipitation;Mann-Kendallmethod;Kolmo-

    gorov-Smirnovtest;Ztest;Ftest;HaiheRiverbasin

    1Introduction

    Withtheglobalwarming,thelocalhydrologicalpatternwillchange,orhasbeenchanging. Received:2009??08?18Accepted:2009?-10-12

    Foundation:NationalBasicResearchProgramofChina,No.2010CB428406;TheKeyKnowledgeInnovationProjectof

    theCAS,No.KZCX2

    YW-126;KeyProjectofNationalNaturalScienceFoundationofChina,No.40730632 Author:ChuJianting,Ph.D,specializedinclimatechange,landsurfaceprocessesandwaterresourcesresearch

    'Correspondingauthor:XiaJun,

    Professor,Email:xi~@igsnrr.ac.cn

    www.scichina.comwww.springerlink.com

    CHUJiantingeta1.:SpatialandtemporalvariabilityofdailyprecipitationinHaiheRiverbasin,19582007249

Oneoftheimportaspectsofthischangeisthespatialandtemporalvariabilityofprecipita

    tion.Inaddition,theabnormalseasonalvariabilityandunevenspatia1.temporaldistribution ofprecipitationareoftenthedirectreasonsfortheextremeflpodanddraughtevents.In China,spatialtemporalvariabilityofprecipRationinwatershedscale,aswellastheimpact ofprecipitationonrunoff,hasbeenpaidmoreandmoreattention.

    TheresearchresultsfromZhangeta1.r2009b)andRen(2007)showthatgreatdiffer

    encesexistinthetemporaltrend&spatialdistributionofprecipitationoverdifferentriver

    basinsinChina.Zheng(2001)studiedthetemperature.precipitationandrunoffoftheYlel

    lowRiverbasinfor1951-1998.TheresultsshowedthatthedischargeintheYellowRiver basinhasbeendecreasingcontinuouslysincethe1950s.however,thisisnottrueasforpre

    cipitation;violenthumanactivityisoneofthemainreasonsforthedecreaseofdischarge. Jiang(2005)analyzedthereasonsfthedischargedecreaseintheTIarimRiverbasinand pointedoutthattheprecipitationinthe1970sislessthaninthe1960s,whiletendstoin

    creaseafterthel970s.Zhangeta1.(2008b)analyzedthespatia1.temporaltrendofextreme precipitationintheYangtzeRiverbasin,andtheyfoundthatthevariabilityofextremepre

    cipitationismilderbeforethe1970s.whilehasobviousincreasingordecreasingtrendafter the1970sdependingontheregionsandseasons;thewetdaysdecreased.whereas.theinten

    sityincreased.especiallyinthemiddleandlowerreaches.Furthermore.allofthesevari

    abilitiescausedthefrequencyoffoodeventsincreasedintheseregions. InChina.theHaiheRiverbasinpossessesspecia1importantstatusinpolitics.economy. andculture.Inrecentyears,thecontradictionbetweenwaterdemandandwatersupplybe

    comesmoreandmoreserious,whichispartlycausedbytheglobalwarmingandsuperfluous groundwaterexploitation.Itisveryurgenttostudyhowtomaintainthesustainableutilities ofwaterresourcesundertheclimatechange.Forthisneed,theresearchhasnotbeendone enoughinHaiheRiverbasinpartlyduetothelimitedaccesstodataandresearchinstru

    mentsfLiu,1999).Inaddition,theexistedstudiesmainlyfocusedontheanalysisofspecific weatherconditioncausingflpodanddroughtevents(Pingeta1.,2003);theanalysisofthe weatherconditionandwatervaportransferovertheHaiheRiverbasin(FanandLiu.1992;

Zhangeta1.,2008);theanalysisofthehistorictrendandforecastofthefuturetrendofrun

    offandwaterresources,etc.(Yuaneta1.,2005;Liueta1.,2004;Shi.1995).Thestudyby Ren(2007)showedthatthedecreasedamountofprecipitationintheHaiheRiverbasinis thelargestamongtenmajorriverbasinsinChina.TheresultsofZhangandWangr2007) gavethesimilarconclusion.Liueta1.(2004)pointedoutthattherelationshipbetweenpre. cipitationandrunoffisdifferentindifferentregionsoftheHaiheRiverbasin.andweneed topaymoreattentiontotheinfluenceofhumanactivities.

    InChina,largescaleprecipitationobservationdataoftencomefromtwosources:the standardmeteorologicalstationsofChinaMeteorologicalAdministrationandtherain gaugesofhydrologicalbureaus.Theformerhaslongermeasuringperiod,anditisgoodto analyzethehistoricaltrendofprecipitation;the1atterhasmorestationsbutshorterobserva

    tionperiods,whichisbettertobeusedtocalibrateandvalidatethehvdrologicalmodels.In addition,precipitationdataaremanagedbythesetwodepartmentsseparately,whichwill impedethedetaileddescriptionoftheirspatialtempora1distribution.

    Recognizingtheaboveconcerns,theobjectiveofthispaperistwofolds:(1)tocompare

    theconsistencyandcorrelativitybetweenthetwoprecipitationdatasets.iefromweather

    JournalofGeographicalSciences

    stations(WS)andfromraingauges(RG),andtoexplorethefeasibilityofextendingpre

    cipitationdataofRGwiththehelpofWSdata;and(2)toobtainthedetailedspa

    tial——

    temporalvariationofprecipitationmakinggooduseoftheadvantagesofthesetwodata sets.Theresearchwillbeusefultoboththehydrologysectorsandthemeteorologicalsectors, andwillprovidesomereferencestotherelatedresearchinotherfiverbasins. 2Dataandmethods

    2.1Data

    TheHaiheRiverbasinislocatedbetween112~-120.Eand35.43.N.andhasanareaof

    about31.8xlO4km2

    ,ofwhich,hillsandplateauoccupynearly60%ofthetotalarea,andthe

    remaining40%belongstotheplainarea.

TothenorthofthewatershedistheYan

    shanMountain.tothewestistheTaihang

    Mountain,andtotheeastisthevastNorth

    ChinaP1ain.Fromnorthtosouth,theHaihe Riverbasincanbedividedintothreeparts: Haihe.LuanheandTuhaimaiiahe.andtheir areasare2.32x10km.4.45x10kmand

    3.18xlkm,respectivelv.Fortheconven.

    ienceofstudy,wedividethewholecatch

    mentinto283sub-basins(Figure1,.

    Inthebasin,thereare30weathersta

    tionswith50years(19582007)ofdata.

    whichareprovidedbytheChinaMeteoro

    logicalAdministration,and248raingauges for10years(19952004)obtainedfromthe

    hydrologicalbureaus.

    2.2Methods

    Inthispaper,5-yearmovingaverageand

    linearregressionmethodswereusedtode

    Figure1Thelocationofweatherstations,raingauges and283subbasinsintheHaiheRiverbasin

    tectthelineartrendpattemoftimeseries(Zhangela1.,2008a,2009a);MannKendall (Zheng,2001;Jiang,2005;Zhangela1.,2008a,2009a;Wang,2008)methodwasusedtodo

    thesignificancetestofthenonlinearchangingpattern;KolmogorovSmirnovtestwasap

    pliedtocheckthedistributionpattern(Xu,2001),andZandFtestwereusedtocheckthe

    equaltyofmeanvalueandvariancebetweenthetwodatasets,respectively.Inaddition,

    someotherstatisticalindicatorswerecalculatedandcomparedwhicharepresentedinthe

    Resultssection.

    Giventhetimeseries{xi},f=1,2.,n,someoftheusedmethodsareexplainedasfol? lows

    2.2.1MannKendalltrendtest

    ThestudyofYueandWang(2002)showthatthehighertheautocorrelationinthetimese

    CHUJiantingeta1.:SpatialandtemporalvariabilityofdailyprecipitationinHaiheRiverbasi

    n,1958-2007251

    ries,thelargertheerroronemayexpectinusingMannKendalltest.Generally,the

    autocorrelationintimeseriesneedstoberemoved,whichisdoneinthefollowingproce

    dure.

    Atfirst.calculatethe1storderautocorrelationcoefficient:

    Cov(~,+1)

    Var(xi)

    c,cxi+I-

    c

    Then,removetheautocorrelationfromtheoriginaltimeseries:

    =PlXix/XiXil(2)一一lL

    Simply,thetransferredseries(),i:1,2.,nisstillnotedas(),i:1,2.,n.

    Second,calculateKendallindicator,variance,,aswellasnormalizedvariableU (Zheng,2001):

    =

    4p

    n(n111(3)

    :

    2(2n

    +5)(41

    9n(n1,

    U=/(5)

    wherePisthenumberofoccurrenceofdualobservationinprecipitationtimeseries. Ucanbeusedtoreflectthemendinhydrologicalormeteorologicaltimeseries.Thelarger

    theIUI,themoreobviousthechangingtrendis.IfU>0,thereisanincreasingtrend,and

viceversa.Giventhesignificancelevel,thecriticalvalue2canbeobtainedfromthe

    standardnormal

    distributiontable;ifIU1>u./2,rejectthehypothesisofnotrend,andsap- posethechangingtrendissignificant.Forexample,given=0.05,then,2u0. 025=1.96;

    ifU>1.96,theincreasingtrendissignificant,ifU<

    1.96,thedecreasingtrendwillbesig

    nifjcant.

    2.2.2KolmogorovSmimov(KS)test

    Theobjectiveofthistestistostatisticallyverifywhetherthedistributionoftheobservation

    dataiSsimilartosomeknowndistilbutionpatterns.Thistestisbasedonthedifferencebe

    tweencumulativefrequencycurveofobservationandtheoreticalfrequencycurveofexpec

    tation(Xu,2001).

    Atfirst,calculatethecumulativefrequencyofthetheoreticaldistribution,i.e.,Fe(x),and

    calculatethecumulativefrequencybasedonthedata,i.e.,(jc): Fo)=k/nf61

    wherekisthenumberofobservations1eSSthanorequalto.and,2iSthetotaJnumberof observations.

    Then.calculatethemaximumdeviationD:

    D=maxWe(x)Fo(x)l(7)

    Giventhehypothesistobetestedas:

    Ho:Fe)=FD(x),Ha:Fe)?,)

    Thehypothesisoftheobservationfollowingthedecideddistributionisrejectedif 252JournalofGeographicalSciences

    JD?D(,z),inwhich,Da(n)isthecriticalvalueandisthesignificancelevel 2.2.3Ztest

    Giventhehypothesistobetestedas:

    H0:fl1--/../2=5,HI:/-/1--,//2?5

    where,6issettobezerofortestingtheequalityofthetwomeansIand2. Calculatethestatistics:

Z一一Y

    /,Jnl

    .

    ,z2

    (8)

    ThehypothesisofthetwomeansareidenticalisrejectedifIzl?z,a2.Givena=0.05,then,

    Za/2=Z0.

    025=1.96.Inwhich,,2,nl,n2arevariancesandlengthsofthetwodatasets, respectively.

    2.2.4Ftest

    Giventhehypothesistobetestedas:

    H.:=,H:?

    Calculatethestatistics:

    2

    F=where>22(9)

    2

    T}lehypothesesofthetwovariancesarethesameisrejectedifF>__F~(nl-1.n2-1).Given thesignificancelevel=0.05,andnl=n2=120,then,(nm1,n2

    1)=Fo.o5(119,119)=1.35will

    befoundfromtheF-distributiontable.

    3Results

    3.1TemporaltrendofprecipitationintheHaiheRiverbasin

    Forthedailyprecipitationof30weatherstationsfrom1958to2007,theareaaverageannual andseasonalprecipitationareobtainedbysimpleaveragingmethod;then,thetemporal trendsofseasonalandannualarea-averageprecipitationarecalculated(Figure21. Itisseenthatinwinter(a),theprecipitationisdecreasingwiththeratearound0.5 mm/10a,whichisnotsignificantatthe5%significancelevel;inspring(b),thereisaminor increase,andtheincreasingtrendisnotsignificantat5%significancelevel;insummer(c), thedecreasingtrendinprecipitationissignificantatthe5%significanceleve1.andtheav

    eragedecreasingrateisabout22mm/10a;inautumn(d),theprecipitationisdecreasingwith

therateabout2.7mm/10a,whichisnotsignificantat5%significanceleve1.Intotal,the

    annualprecipitation(e)isdecreasingwitharateof23mrn/lOa,andthistrendissignificant

    at5%significanceleve1.

    3.2SpatialandtemporalrelationshipofprecipitationintheHaiheRiverbasin

    Asmentionedabove,thelongtermprecipitationdatafromWScandescribethetemporal

    changingtrendofprecipitationinthewholeriverbasin,however,itcannotpicturethede

    CHUJiantingeta1.:SpatialandtemporalvariabilityofdailyprecipitationinHaiheRiverbasi

    n,1958-2007253

    g40

    30

    e-

    20

    lo

    o

    :一?46x+~Observationp=048345-yearmovigaverag.

    .

    \'me

    n^^..r,A,.

    \7.

    V

    .

    V\.

    1958l963l968l973l978

    170

    120

    0

    7O

    2O

    1983l9881993l9982003

    Year

    (a1Winter |b)v:0.223lx+64.795_'uuu"

    .

    s^/\.^,?a^ve?rage'

    ..

    

    ..

    V.

    1958l963l9681973l9781983I988l99319982003

    Year

    fbSpring 200

    150

    

    100

    5o

    0

    Year

    (c)Summer ^Zb/~4x

    //

    .

    0s

    ear

    crv

    moVallollaVe

    ,

    

Report this document

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