城市暴雨內(nèi)澇情景模擬與災(zāi)害風(fēng)險(xiǎn)評(píng)估-英文-

DI:10.1007/442-01-084-7?201ScDI:10.1007/442-01-084-7?201SciencePressSpringe-erlagCommunity-basedscenariomodellinganddisasterriskassessmentofurbanrainstormwaterloggingYINZhan’e1,YINJie2,*XUShiyuan2,WENJiahong11.GeographyDept.ofShanghaiNormalUniversit,Shanghai200234,China;2.GeographyDept.ofEastChinaNormalUniversit,Shanghai200062,ChinaAbstract:Scenariomodellingandtheriskassessmentofnaturaldisastersisoneoftheho-sposindisasterresearch.Howeve,upuntilno,urbannaturaldisasterriskassessmentslackcommonproceduresandprogrammes.Thisaperselecsrainstormwaterloggingasadisastertoresearch,whichisoneofthemostfrequentlyoccurringhazardsformostcitiesinChina.Asanexample,weusedasmall-scaleintegratedmethodologytoassessrisksrelatingtorainstormwaterlogginghazardsintheJing’anDistrictofShanghai.Basedonthebasicconceptofdisasterrisk,thisaperappliesscenariomodellingtoexpresstheriskofsmall-scaleurbanrainstormwaterloggingdisastersindiferentreturnperiods.Throughthisanalysisofvulnerabilityandexposure,wesimulatediferentdisasterscenariosandproposeacomprehensiveanalysismethodandprocedureforsmall-scaleurbanstormwaterloggingdisasterriskassessments.Agrid-basedGeographicalInformationSystem(GIS)approach,includinganurbanterrainmodel,anurbanrainfallmodelandanurbandrainagemodel,wasappliedtosimulateinundationareaanddepth.tage-damagecurvesforresidentialbuildingsandcontenswerethengeneratedbythelossdaaofwaterloggingfromfieldsurveys,whichwerefurtherappliedtoanalysevulnerabilit,exposureandlossassessment.Finall,theex-ceedanceprobabilitycurvefordisasterdamagewasconstructedusingthedamageofeachsimulatedeventandtherespectiveexceedanceprobabilities.Aframeworkwasalsodevel-opedforcouplingthewaterloggingriskwiththeriskplanningandmanagementthroughtheexceedanceprobabilitycurveandannualaveragewaterloggingloss.Thisisanewexplora-tionforsmall-scaleurbannaturaldisasterscenariosimulationandriskassessment.Keywords:scenariomodelling;small-scale;rainstormwaterlogging;disasterriskassessment;Shanghai1Intoductionithglobalcliatechangeandrapidurbanization,theintensityandfrequencyofurbannaturaldisastersandassociatedlossesareincreasing(Xuetal.,2006;Shi,2006;ZhouandReceived:2010-0-20Accepted:2010-1-28Foundation:NationalNatureScienceFoundationfChina,No.410714;No.4073526;KeySubjectDevelopingProjectbyShanghaiMuncipalEducationComission,N.J50402;ScienceandechnologyCoissionofShag-haiMunicipalit,No.082451000;LeadingAcadeicDisciplineProjectofShanghaiNoralUniversit,No.DZL809to:inZhan’e(1963–),Ph.DandAssociateProfesso,specializedinnaturalhazardsandreotesensing.E-ail:zhanein126*Corespondingauthor:XuShiyuan,Professor,E-ail:syxu@HYPERLINK://geogsci/eogscispringrlink/content/1009-67Xuan,1993).aterloggingisoneofthemostseriousnaturauan,1993).aterloggingisoneofthemostseriousnaturaldisasterphenoena,notonlyinChinesecitiesbutalsoaroundtheworld,causingconsiderablepersonalinjuryandprop-ertydaage.Forexaple,morethan200roadsand50,000buildingsandhouseswerein-undatedbywaterloggingfromyphoonMatsain2005inShanghaiandthedirecteconoiclosswasaboutRMB1358millionyuan.Thus,itisimportanttotakeintoaccountdisastermanageenttoreduceeconoicloss,socialipactandcasualtiesfromnaturalhazards(Suetal.,2005;Hays,1991).Indisasteranageent,riskassessentshavebecoearesearchhotspot.Therearediferentethodstoconductlage-scaledisasterriskassessments.How-eve,fewethodologiesareavailableforsmall-scaledisasterriskassessments.Theainobjectivesofthisstudyaretoprovideafraeworkandanalternativemethodforsmall-scaleriskassessmentsofurbanwaterlogging.ithinthefraework,theJing’anDistrict(anur-bancentreandanareainShanghaipronetowaterlogging)istakenasacasestudytode-onstratetheproposedethodolog.2Materialsandmethods2.1tudyaeaanddatasetsTheJing’anDistrict,withitscentrelocatedat31.13oNand121.26oE,isoneofthe19ad-ministrativedistrictsinShanghai,China(Figure1).ithanareaof7.62k2,itisco-prisedoffiveblocks,withapopulationreaching248,400andadensityof32,598peo-ple/k2in2008(ShanghaiMunicipalStatisticsBureau,2009).Mostoftheregionisveryflatandlowlying,withanaveragealtitudeof3maboveWusongDatu.Duetoadecreaseinthenuberofriversinthecitycentreandasubsequentweakeningofexcess-waterstor-agecapacity(Chengetal.,2007),surfacerunoffhasincreased.Asoneofthetraditionalur-bancentres,thisareahasthetypicalcharacteristicsofanurbanlandscape,suchascrowdedbuildings,heavytraffic,an“urbanheatisland”and“urbanrainisland”,thusmakingtheFigue1Locationofthestudyarea(1-Jing’an;2-Changning;3-Putuo;4-Zhabei;5-Hongkou;6-Yangpu;7-Huangpu;8-Luwan;9-Xuhui)276JournalofGeogrphicalSciencesDistrictorevulnera276JournalofGeogrphicalSciencesDistrictorevulnerabletowaterlogging.ThedatasourcesusedinthisstudyincludetheadinistrativeboundarydataofShanghaiin2006,thecensusdataofShanghaiin2006,landuseandlandcovermapsofShanghaiin2006,thetopographiccontours(0.5mcontourintervals)datarelativetotheusongerticlDatumofShanghaiin2005andthebuildingfootprintsoftheJing’anDistrictobtainedbymanualvisualinterpretingfromaerialphotographs(0.36mhorizontalresolution)in2006.aterlogginglossdatawasobtainedfromin-situinvestigationonahouse-by-housebasisusingtraditionalpaperquestionnaires.2.2 MethodsRiskcanbedefinedasthepossibilityofexpectedlosses(e.g.,casualties,injuries,propertydaageanddisruptionofeconoicactivitiesorenvironentaldegradation)duetohazard-ouseventsinagivenareaduringaspecificreferenceperiod(UNDRO,1991;Helm,1996).Itcanbeexpressedasfollows:R=H∩V∩E(1)whereRisrisk;Hishazard,theprobabilityofapotentiallydaagingphysicalevent,phe-noenonorhuanactivitythataycauseanexpectedlosswithinadeterinedperiodandregion;Visvulnerabilit,thedegreeoflosstoagiveneleentatriskorasetofeleentsatriskresultingfromtheoccurrenceofanaturalphenomenonofagivenagnitude;Eisex-,eeetatkngi,ao,ptyrrunaeAsamajorandfrequentlyoccurringnaturalhazard,waterloggingriskcanalsobeas-sessedusingtheaboveequation.Sofa,commonexperientalhydrologicalodelsandhy-draulicsnuericalsiulationshaveonlybeenusedtoanalysethecharacteristicsofwater-logginghazard.Theexperimentalhydrologicalodels,suchasSORMandSCS(U.S.SoilConservationService,1972)models(angandu,2009)aresiplebutitcanhardlyex-pressthedynaicsandthedetailedspatialinforation.Conversel,the1Dand2Dhydrau-licsnuericalsiulationsareusefulindepictingthespatialandteporaldistribution(ZhengandHu,2003;Congetal.,2006;Qiuetal.,2000;Lietal.,2009).Howeve,thesemethodsrequireuchspecificdata,whichcannotbeacquiredeasilyanddeandassivedataprocessing;thus,itisdificulttousethemforsmall-scaleassessments(enandAkan,1999).Otherthanhazardassessmentdata,vulnerabilityinformationreainsliitedinChina.Lessattentionhasbeenpaidtotheconstructionofvulnerabilityassessmentodelsandtheinvestigationofdisasterdaage.Basedonpreviousstudies(Wangetal.,2004;Zhaoetal.,2004;Jonkanetal.,2008;Duttaetal.,2003),acoprehensiveurbanwaterloggingriskassessmentmodelwasdevel-opedtosimulatedifferentinundatedscenariosandbuildvulnerabilitycurves(orstage-daagecurves/lossfunction);wethencalculatedtheexpectedloss.Consideringthecomplexityofunderlyingsurfaceandhuman-induceddisturbances,grid-basedscenarioanalyticalapproaches(includingurbanterrainodelling,urbanrainfallodellingandurbandrainageodelling)wereeployedtoderivethehazardouspotentialinhazardanalysis.SeveralvulnerabilitycurveswerebuiltfordifferenthuanactivitiesbasedonfieldsurveysafterwaterlogginginShanghai.Huanactivitiesthataresubegedbeneathwaterloggingcanberevealedinexposureanalysis.Finall,theriskcanbeexpressedasanexceedanceFigue2FrameworkofrainstormwaterloggingriskassessmentbasedFigue2FrameworkofrainstormwaterloggingriskassessmentbasedonaGIS-gridprobabilitycurve,ariskcurveandanaverageannualwaterloggingloss.Figure2exhibitsthefraeworkfortheurbanwaterloggingriskassessment.3HazardanalysisAsthefirststepofriskassessment,hazardanalysisistoidentifysomeparametersofhaz-ardouspotentialindiferentreturnperiods(Plate,2002).Inthispape,waterdepthwasper-hapsthebestindicatorfortheipactfromwaterlogging.Basedonpreviousstudies,theur-banterrainmodel,theurbanrainfall-runofodel,theurbandrainageodelandthe“hy-drology”toolinthe“SpatialAnalyst”moduleofArcGISwereintegratedwithinacopre-hensiveanalysisproceduretodetectthedepthofinundatedwate.3.1 CatchmentaeaAtopographicdatasetintheformofadigitalelevationodel(DEM)isthebasisforhydro-logicandfloodinundationmodelling.Theelevationinthetopographiccontourswasinter-polatedtogenerateaDEMwithagridcellresolutionof30musingtheArcGIS9.2software.Consideringthatthedirectionofflowisafectedbygravit,theflowdirectionapwasgeneratedusingtheShanghaiDEM;thecatchentswerethenobtainedusingtheGISap-plication.Somecatchents,whichwereinvolvedintheJing’anDistrict,weremegedintoawatershed.Furtherore,thewatershedboundarywasmodifiedaccordingtoactualurbanterraincharacteristics,suchasriverbanksandroads.3.2 UrbanterrainmodelBecauseoftheconstructionofbuildingsandotherfabricatedobjects,thereisalagedifer-encebetweenurbanterrainandnaturalgroundsurface.Thespatialdistributionofwaterwillbeipactedbythebuildingsthatresistwaterflow(Quanetal.,2010).AsLightDetectionandRanging(LiDAR)dataareunavailableforthestudyarea,theoriginalDEMwasrevisedbytheurbanterrainodelwiththeadditionofgroundclearanceofthebuildings’groundfloors(Figure3).Accordingtothecodeforthedesignofcivilbuildings(JGJ37-87)andat-278JournalofGeogrphicalSciencestributeinformationofbuildingsintheJing’anDistrict,whichwas278JournalofGeogrphicalSciencestributeinformationofbuildingsintheJing’anDistrict,whichwascollectedthroughexten-sivefieldworkafterwaterloggingeventsinAugust26,2008andSepteber20,2008,theclassificationofbuildingsandattributedataintheJing’anDistrictarelistedinTable1.Figue3ModificationofurbanDEMbybuildingsable1ClassificationofbuildingsandattributedataintheJing’anDistrictGroundclearance(c)Replacementvalueofbuilding(yuan/2)Replacementvalueofbuildingcontents(yuan/2)ypesResidentialbuildings35(floornu-ber>9,60)Class1Gardenhouse,villa80001000Class2High-riseapartmentbuildingandbuildings35(floornu-4000500ofresidentialnewvillagesbefore1988ber>9,60)Class2NComodityhouseandbuildingsofresiden-35(floornu-5000800tialcomunityafter1988ber>9,60)Class3Li-Nong(lanesandalleys)residentialbuild-102500300ingsandclass3buildingsofeployeesClass4Shacks510001503.3UrbanrainfallrunoffmodelRainfallisthesourceofwaterloggingandthekeyinputsforhazardanalysis.Inthispape,theurbanrainstormintensityforulaofShanghai,proposedbytheShanghaiMunicipalEn-gineeringDesignGeneralInstitute,wastakenastheurbanrainfallodeltocalculatedi-ferentintensitiesofrainfall(ShanghaiMunicipalEngineeringDesignGeneralInstitute,2000).Theequationcanbewrittenasfollows:q=1995.84(P0.30–0.42)/(t+10+7lgP)0.82+0.07lgP(2)whereqistherainfallintensit,Pisthereturnperiodofrainfallandtisthedurationofrain-fall.TheSCSodel,developedbytheU.S.SoilConservationin1972,wasusedastheurbanrunofodeltodeterineurbansurfacerunof(U.S.SoilConservationService,1972).Ithasbeenappliedwidelyandrecognizedasapowerfulandcost-efectivetoolfordetectingandanalysingtheaountofdirectrunofandthepotentialrunoflossesfromadsorptionandinterruptionofplants,evapotranspiration,infiltrationandsoon(Richardetal.,2006,Shietal.,2001).Theforulaisasfollows:Q=(T–Ia)2/(T+S–Ia)SisderivedfromanequationexpressediQ=(T–Ia)2/(T+S–Ia)Sisderivedfromanequationexpressedintermsofthecurvenuber(CN):S=25,400/CN–254(3)(4)whereQisthedirectrunof,Tisthetotalrainfall,Iaistheinitialabstraction,Sisthepoten-tialaxiumretentionandCNisderivedfromthetablesgivenintheNationalEngineeringHandbook(Section-4(NEH-4)forcatchentscharacteristics).Inthisstud,theIa(0.05S)andCNvaluesofShanghai,proposedbyHeetal.(2003),wereusedtosiulateurbansur-facerunof.Thedrainagesystemsmustalsobeconsideredforurbanareas.Duetothelackofdetailedsewe-drainagesysteminformation,itwasassuedthatthedrainagesystemwasevenlydistributedinthestudyregion.Atpresent,thedrainagesystemofthecitycentreofShanghaiprovidesprotectiontoareturnperiodofoneearforthedesignrainfallintensity(36mm/h).Therefore,thefinalaountofwaterloggingcanbecalculatedbyEquation(5):W=(Q–D)×S'(5)whereWistheaountofwaterlogging,DistheaountofdrainageandS'isthecatchentarea.3.4 ScenariossimulationofurbanwaterloggingRainstormwaterloggingbelongstothe“non-sourcesubmeged”case.Inthiscase,thefloodroutingodelsspecifythatagridcellisfloodedonlyifitselevationisbelowfloodlevel.Basedontheaboveanalysis,inundationextentanddepthweredeterinedbythe“equal-volumeethod”,whichassuestheaountofwaterloggingisequivalenttofloodvolue(Zhang,1995;LiuandLiu,2001;GuoandLong,2002).Theregionsbelowahypo-theticalfloodlevelweredeterinedusingthefloodroutingodel,andthentheaccommo-datedfloodvoluewascalculated.Oncethefloodvoluewasequaltotheaountofwa-terloggingintheroutingmodel,thehypotheticalwaterlevelintherespectiveregionswasdeterined.oassessthewaterloggingriskforthestudyarea,rainstormscenarioswiththedurationofonehourandareturnperiodof5,10,20,50,100,200,500and1000yearsweresiulatedforthishazardanalysis,tocoveracopletesetofprobablesituations.Thesiu-lationresultsandsoeinundationscenariosaredisplayedinFigure4andable2.4ulnerabilityanalysisulnerabilityanalysesareusedtodetecttherelationshipbetweenloss(orlossrate)andsev-eralhazardfactors,suchaswaterdepth,flowvelocityandduration(Arnell,1999;Berningetal.,2000,2001;Goldan,1997;Kangetal.,2006;Smith,1994;Maaskantetal.,2009).Flowvelocityanddurationwerenottakenintoconsiderationinthisstud,asthewaterlog-ginginShanghaiwascategorizedasflashwaterloggingwithlowflowspeed.Therefore,astage-daagecurves,whichdescribetherelationshipbetweenloss/daageandwaterdepth,werebuiltforresidentialbuildingsandtheircontents.Peoplesvulnerabilitywasestiatedusingtheortalityfunction,poposedbyJonkan(2003).Duetoalackofreliabledataandlossrecords,otherhuanactivitieswerenottakenintoconsiderationinthisstud.Thedatausedtoderivethestage-daagecurveswerecollectedfromseveralfieldsurveysinthestudyarea.Althoughthismethodpresentsfewliitations,itisstillacomonlyusedap-proachtoevaluateurbanlossesafterhazardanalysis.280JournalofGeogrphicalSciencesFigue4Inundationscenariosofdiferentreturnperiodsofth280JournalofGeogrphicalSciencesFigue4InundationscenariosofdiferentreturnperiodsoftheJing’anDistrictable2DepthandfloodinundatedareaofdiferentreturnperiodsintheJing’anDistrictReturnperiodsParaeters51020501002005001000Rainstormintensity(m/h)aterlevel()Inundationarea(k2)Maxiumwaterdepth()582.603.10.10672.673.50.17772.734.00.23912.804.40.301012.854.80.35142.895.10.391302.955.50.451443.007.20.50Inordertoestiatedaagetoresidentialbuildings,buildingconstructiontypesandtheirreplaceentvaluesweretakenintoaccount.obuildingconstructiontypesprevailintheJing’anDistrict:reinforcedconcreteandasonry(CM)andaixtureofconcreteblocksandwood(CW).Accordingtothecitysurbanplan,CWwillbereovedbefore2010.v-eragereplacementvalueswereobtainedbyaveragingconstructioncostsfordiferentsocio-economicclasses(able1).Therefore,thecurvewasplottedonlyfortheCMbuild-ings(Figure5).odeterinethevulnerabilitytowaterloggingforresidentialbuildingcontents,alistofresidentialbuildingscontentsandcontentreplaceentvaluesneededtobeidentified.Alistofarticlesinresidentialbuildingswasderivedfromthefieldsurveydataandtheinterviewswithsoeinhabitants.ThecontentreplaceentvalueswereobtainedfromshopsinShang-hai.AcurveforbuildingscontentswasdevelopedasshowninFigure6.5ExposueanalysisExposureanalysisisusedtodetecttheeconoicactivitiesadverselyafectedbywaterlog-Figue5tage-damagecurveforresidentialbuildingsFigue6tage-damagecurveFigue5tage-damagecurveforresidentialbuildingsFigue6tage-damagecurveforresidentialbuildingcontentsgingusinghazardinforation,stage-damagecurvesandeleentsatrisk.Exposureanalysisalsoindicatesthedegreetowhicheleentsatriskareexposedtodiferentrainstormwater-loggingintensities.Fortheresidentialbuildingsandpeopleinthisstud,thespatialexpo-sureinforationwasdepictedinGISbyoverlayingthewaterloggingextentapwiththeresidentialbuildingfootprintandpopulationdensitymapsoftheJing’anDistrict.Theresi-dentialbuildingsandpeoplecoveredbythewaterloggingwerecategorizedastheexposure.Fortheresidentialbuildingcontents,thespatialexposureinforationwasobtainedbysub-tractingthebuildinggroundlevelfromthewaterlevel.aterdepthsofbuildingcontentsgreaterthan0wereassignedtheeposure,aspresentedinFigure7.6aterloggingdamageandriskassessmentaterloggingdaageassessmentisaniportantstepincalculatingexpectedlossesforresidentialbuildings,residentialbuildingcontentsandpeople.Thiswasdonebyintegratinginforationderivedfromtheabovesections.Accordingtoourcalculations,nocasualtieswereprojectedintheJing’anDistrict.odisplayclearlythespatialinforationofeachbuilding,agridof2×2mwasusedfortheriskassessentinthisstud.Thetotaldaagevaluescanbeaggregatedcell-by-cellfordiferentreturnperiods(Figure8).Theriskofeachrainstorm-waterloggingscenariowascalculatedbymultiplyingtheexpectedlossesofthatscenariowithitsoccurrenceprobabilit.Onthisbasis,theriskcurvecanbeconstructedus-ingtheriskofeacheventanditsrespectiveoccurrenceprobabilities(Figure9).Theaverageannualwaterloggingloss(AWL)canbedefinedastheareaundertheriskcurveandcanbecalculatedbyEquation(6)(Hardison,1972).AWL=∫xf(x)dxwherexistheoccurrenceprobabilityofawaterloggingeventandf(x)isthelossofx.(6)TheAWLoftheJing’anDistrictwasaboutRMB3.42illionyuanandcanbeusedasabasisforcost-benefitanalysis.Inaddition,theanalysis-of-riskcurveindicatedthatanoc-currenceprobabilitylessthan0.02wouldbeadisastrousriskbecausetheaximumwaterdepthsofthesescenarioswouldexceed25c,whichwouldresultinaseriousipactonurbanfunction,inrelationtotheclassificationstandardofurbanrainstormwaterlogging.Theresultswillenablepolicymakersorriskmanagerstoidentifywheretherearehigherrisksandwhichareasneedmoreriskanageentattention.282JournalofGeogrphicalSciencesFigue7Exposureofbuildin282JournalofGeogrphicalSciencesFigue7ExposureofbuildingcontentsofdiferetreturnperiodsintheJing’anDistrictFigue8LossmapsofdiferentreturnperiodsintheJing’anDistrictin2008Figue9RiskcurveofrainstormwaterloggingdamageintheJing’anDistrict7DiscussionandconclusionsThisstudypresentsacombinedethodologyofhydrologicalodellingandGISspatialanalysistoanalysequantitativelytheriskofrainstormwaterlogginginasall-scaleurbanarea.Scenarioanalysiswasusedtoidentifythedepthandextentofwaterloggingusingacoprehensiveanalysisprocedure.Anewurbanterrainodelwasproposedasanalterna-tiveforLiDAR.Undereighthypotheticprecipitationscenarios,thehazardouspotentialintheJing’anDistrictwassimulatedusingGIS.tage-daagecurvesaccordingtothelocalconditionsweredevelopedfortheresidentialbuildingsandtheresidentialbuildingcontentsusingtheinformationcollectedthroughinterviews,whichwerefurtherappliedtoanalysethevulnerabilit,exposureanddaageofwaterlogging.Finall,thelossmapsofdiferentreturnperiods,theriskcurveandAWLweretakenastheexpressionofwaterloggingriskofthestudyarea.Theresultsshowthattheaxiuminundationdepthwas50cm,andtheAWLwasaboutRMB3.42illionyuan.AdetailedriskassessAWLwasaboutRMB3.42illionyuan.Adetailedriskassessmentmaycontributesignificantlytoalleviatethenegativeefectsofnaturalhazards,andtohelppolicyakersandotherconcernedstakeholdersdevelopappro-priateitigationstrategiesandanagementchoicesbeforesuchproblemsbecoeirre-versible.Howeve,tobemorecoplete,thisstudycouldbefurtherimprovedinthefollow-ingaspects:(1)oreaccurateandcoprehensivedatawouldprovideoreaccurateresults,(2)hydrodynaicodelcouldbeintroducedintotheinundationsiulationandanalysisand(3)lossesfromeco-environentaldamageandindirectlossesshouldbeconsidered.RefeencesArnellN,1999.Expectedannualdamageanduncertaintiesinfloodfrequencyestimation.JournalofaterResoucesPlanningandManagement,15:94–107.BerningC,DuPlessisLA,iljoenM,2001.Lossfunctionforstructuralfloodmitigationmeasures.aterSA,27(1):35–38.BerningC,iljoenM,DuPlessisLA,2000.Lossfunctionforsuga-cane:Depthanddurationofinundationasdeterminatesofextentofflooddamage.aterSA,26(4):527–530.ChengJiang,angKai,ZhaoJunetal.,2007.ariationofriversystemincenterdistrictofShanghaianditsim-pactfactorsduringthelastonehundredyears.ScientiaGeographicaSinica,27(1):85–91.(inChinese)CongXiangyu,NiGuangheng,HuiShiboetal.,2006.SimulativeanalysisonstormfloodintypicalurbanregionofBeijingbasedonSWMM.aterResoucesandHydopowerEngineering,(4):64–67.(inChinese)DuttaD,HerathS,MusiakeK,2003.Amathematicalmodelforfloodlossestimation.JournalofHydology,277:24–49.GoldmanD,1997.Estimatingexpectedannualdamageforleveeretrofits.JournalofaterResoucesPlanningandManagement,123(2):89–94.GuoLihua,Longi,2002.AnalysisoffloodsubmegingbasedonDEM.BulletinofSurveyingandMapping,(1):25–30.(inChinese)HardisonCH,JenningsME,1972.Biasincomputedfloodrisk.JournaloftheHydraulicsDivision,98(3):415–427.HaysW,1991.HazardandriskassessmentintheUnitedtates.Episodes,14(1):7–12.HeBaogen,ChenChungen,ZhouNaisheng,2003.Urbanizedarearunofcoeficientanditsapplication.ShanghaiEnvionmentalSciences,22(7):472–475.(inChinese)Helm,1996.Integratedriskmanagementfornaturalandtechnologicaldisasters.ephra,15(1):4–13.JonkmanSN,2007.Lossoflifeestimationinfloodriskassessment[D].DelftUniversit.JonkmanSN,BockarjovaM,KokMetal.,2008.IntegratedhydrodynamicandeconomicmodelingofflooddamageintheNetherlands.EcologicalEconomics,66:77–90.KangXiangwu,uShaohong,DaiErfuetal.,2006.Lage-scalefloodlossesandtheimpactpre-assessment.ChineseScienceBulletin,51(suppl.):155–164.(inChinese)Lieifeng,ChenQiuwen,MaoJingqiao,2009.tudyonfloodriskmodelingofBeijingOlympicillage.Chi-neseScienceBulletin,54(3):321–328.(inChinese)LiuRenyi,LiuNan,2001.AGISbasedmodelforcalculatingoffloodarea.ActaGeographicaSinica,56(1):1–6.(inChinese)MaaskantBob,JonkmanSebastiaanN,BouwerLaurensM,2009.Futureriskofflooding:AnanalysisofchangesinpotentiallossoflifeinSouthHolland(TheNetherlands).EnvionmentalScience&Policy,12:157–169.MinistryofUrbanandRuralConstructionandEnvironmentalProtectionofthePeoplesRepublicofChina,1987.JGJ37-87:CodeforDesignofCivilBuildings.(inChinese)PlateEJ,2002.Floodriskandfloodmanagement.JournalofHydolog,267:2–1.QiuJingwei,LiNa,ChengXiaotaoetal.,2000.ThesimulationsystemforheavyrainfallinianjinCit.Journal284Jou