非点源污染模型AnnAGNPS在三峡库区林农复合小流域模拟效果评定

AnnAGNPS1,1,2,1,1,3(1.,,100091;2.,471003;3.,443600):AnnAGNPS(AnnualizedAGriculturalNonPointSourceModel),20032004,(R2)Nash2Sutcliffe(E)(VE).,,VE510%(R2=0193,p0105),VE617%(R2=0190,p0105);,,VE1511%(R2=0163,p0105),VE2617%(R2=0159,p0105);,,,.R20168(p0105),(R2=0165,p0105)..AnnAGNPS.:AnnAGNPS;;;;;:X522:A:025023301(2009)1022872207:2008211207;:2009202217:(2006BAD03A13,2006BAD03A07):(1966),,,,,E2mail:hzlin66@1631comNonpointSourcePollutionModel,AnnAGNPS,AssessmentforaMixedForestedWatershedinThreeGorgesReservoirAreaHUANGZhi2lin1,TIANYao2wu1,2,XIAOWen2fa1,ZENGLi2xiong1,MADe2ju3(1.KeyLaboratoryofForestEcologyandEnvironment,StateForestryAdministrationChina,ResearchInstituteofForestEcology,EnvironmentandProtection,ChineseAcademyofForestry,Beijing100091,China;2.CollegeofForestry,HenanUniversityofScienceandTechnology,Luoyang471003,China;31ForestryBureauofZiguiCounty,Yichang443600,China)Abstract:Watershedmodelsprovideacost2effectiveandefficientmeansofestimatingthepollutantloadingsenteringsurfacewaters,especiallywhencombinedwithtraditionalwaterqualitysamplingandanalyses.Buttherehaveoftenbeenquestionsabouttheaccuracyorcertaintyofmodelsandtheirpredictions.ThemaingoalofthisstudywastoevaluatetheperformanceofAnnAGNPS(AnnualizedAGriculturalNonPointSource)PollutionModel,insimulatingrunoff,sedimentloadingandnutrientloadingsunderThreeGorgesReservoirarea.MostofmodelinputparametersweresourcedfromZiguiForestEcologyStationinThreeGorgesReservoirarea,StateForestryAdministration.Datayear2003wasusedforcalibrationwhiledatayear2004wasusedforvalidationofthemodel.Thewholeevaluationconsistedofdeterminingthecoefficientofdetermination(R2),Nash2Sutcliffecoefficientofefficiency(E),andthepercentagevolumeerror(VE).Resultsshowedthatthemodelpredictedthedailyrunoffvolumewithintherangeofacceptableaccuracy.Therunoffonadailybasiswasunderpredictedby510%withR2of0193(p0105)duringcalibrationandunderpredictedby617%withR2of0190(p0105)duringvalidation.Butsedimentloadingwasabletoproduceamoderateresult.Themodelunderpredictedtheevent2basedsedimentloadingby1511%withR2of0163(p0105)duringcalibrationand2617%withR2of0159(p0105)duringvalidation.Fortheeventsofsmallmagnitude,themodelgenerallyoverpredictedsedimentloading,whiletheoppositewastrueforlargerevents.NitrogenloadingpredictionwasslightlybetterwithR2=0168(p0105),andphosphorusloadingperformancewasslightlypoorwithR2=0165(p0105).Ingeneral,themodelperformswellinsimulatingrunoffcomparetosedimentloadingandnutrientloadings,andasawatershedmanagementtoolsitcanbeusedforThreeGorgesReservoirareaconditionsthatwithmixedtypesoflandusesandsteepslopes.Keywords:AnnAGNPSmodel;runoff;sedimentloading;nutrientloading;ThreeGorgesReservoirarea;performance(nonpointsource,NPS).NPSANSWERS[1]CREAMS[2]AGNPS[3,4]SWAT[5]AnnAGNPS[6],AnnAGNPS(AnnualizedAGriculturalNonPointSource)ArcViewGIS31x,.Jennifer[7]3010200910ENVIRONMENTALSCIENCEVol.30,No.10Oct.,2009©1994-2010ChinaAcademicJournalElectronicPublishingHouse.Allrightsreserved.[8]CurrencyCreek,Tsou[9]BedrockCreek,,,,NPS.AnnAGNPS,[10][11].AnnAGNPS,,,(),.,.NPS[12,13].,,(),,AnnAGNPS,(bestmanagementpractices,BMPs).1111,11053271105450,305121305134,14414hm2,(1),,1439mm,,(5868%).,,7215%.,1400m,,;;,.4518hm25611hm2619hm22018hm2.,203.112AnnAGNPS1Fig.1LocationofthestudyareaAnnAGNPS(USDA2ARSNRCS),AnnAGNPS410(2007),AGNPS,ArcView31x.3.[(1)](SCS2CN)[6][(2)(3)].RUSLE[(4)];EPIC[14],,378210:AnnAGNPS©1994-2010ChinaAcademicJournalElectronicPublishingHouse.Allrightsreserved.=SMt+WIt+Qt+PERCt+ETt+Qlat+QtileZ(1),SMtSMt+1(%),WIt(mm),Qt(mm),PERCt(mm),ETt(mm),Qlat(mm),Qtile(mm),Z(mm),t(d).Q=WI-012S2WI+018S(2)S=254100CN-1(3),Q(mm),WI(mm),S,CN.A=RKLHCP(4),A[t(hm2a)-1],R,K,L,H,C,P.113(DEM)510m110000,,ArcViewGIS312[15]510mDEM.50,,DEM;300,GPS,DEM[16].AnnAGNPSCSA(CriticalSourceArea)MSCL(MinimumSourceChannelLength).,,CSAMSCL2hm230m,66,(2).2Fig.2HeigouwatersheddiscretizationunitsdelineationsUNIXGEMDayclim.inp.,soils.shpfields.shp;;,InpEdit.AGNPSPArvViewInterface,AnnAGNPS.inpDayclim.inp,.1.1Table1DatasourcesfortheHeigouwatershedGIS110000110000DEMGIS110000pH110000114Nash2SutcliffeE[17][(5)]VE[(6)]R2[(7)]3.E=110-ni=1Mi-Pi2ni=1Mi-ŠM(5)VE=ni=1Mi-Pini=1Mi(6)R2=ni=1Mi-ŠMMi-Pi2ni=1Mi-ŠM22ni=1Pi-…P22(7)478230©1994-2010ChinaAcademicJournalElectronicPublishingHouse.Allrightsreserved.(),Mii,ŠM,Pii,…P.E,,11[18],E,,E016,[19],,E1.,E[20];VE,,,.Chiew[19]VE15%;R2,.115,.-4%4%9RCN,AnnAGNPS.inp2003Dayclim.inp,PLModel,,23,[8],2%RCN,1%RCN,,VE15%[19].,.,(rootmass)(canopycover)(Manningsroughnesscoefficient)[21].,PEST,PLModel,,[8],.116,2004AnnAGNPS.inp,2004Dayclim.inp,PLModel,2004,49,R2VEE3.2211AnnAGNPS.EVER2.,,11(3),VE-510%,15%.R2=0193(p0105),E=0187,016.32003Fig.3MeasuredvspredicteddailyrunofffromHeigouwatershedduringcalibrationin200311(4),VE-1511%,,R2=0163(p0105),E=0177016.AnnAGNPS.42003Fig.4MeasuredvspredicteddailysedimentloadingfromHeigouwatershedduringcalibrationin2003212AnnAGNPS22004.212112004492132105m3,2145105m3,VE-513%.578210:AnnAGNPS©1994-2010ChinaAca