基于RSGIS的水电开发生态环境时空影响定量分析研究

华中科技大学博士学位论文基于RS/GIS的水电开发生态环境时空影响定量分析研究姓名：李文波申请学位级别：博士专业：空间信息科学与技术指导教师：张勇传;张秋文20080606I――1LandsatTM/ETM+GIS2RS/GIS1987465.31km220005102.25km22004112.74km2198720041772.673198723.93II23.1825.432200015.7521.4445.14200411.6620.0455.844GIS//21.7729.241.6013.98RS/GIS//IIIAbstractWaterresourceisoneoftheirreplaceablestrategicresourcesforhumansurvivalandsocialdevelopment,andthehydropowercomingfromwaterresourcesisonekindofcleanenergy.Thedevelopmentandutilizationofhydropowercanreplacesomethermalpowersoastoreducingenvironmentalpollution.Tomakefulluseofwaterresource,theexploitationofcascadehydropowerinariverbasinisaninternationaltendencyandalsothefundamentalhydropowerenergypolicyinChina.Thedevelopmentofcascadehydropowercanbringenormoussocialandeconomicbenefits,butitalsomayhavesomenegativeecologicalenvironmentimpacts(EEI).Eco-environmentalimpactsassessment(EIA)isoneofthemostimportantmeasureswhichcanminimizethenegativeeco-environmentalimpactsinducedbycascadehydropowerdevelopment.However,theEIAisobviouslyregionalandspatial,andthetraditionalevaluationmeansbasedonthetraditionalgroundmonitoringnetworkaretooinefficiencytomeettheneedsofmanagementanddecisionofeco-environment.Whereasthespatialinformationtechnologies,especiallyremotesensingtechnologyandGeographicInformationSystem(GIS),owingtotheirmacroeconomic,fastandeconomiccharacteristics,canbepowerfultechnologicalmeansforeco-environmentevaluation.Sofar,thespatialinformationtechnologieshavebecomeimportantresearchmeansofglobalandregional(basin)eco-environment.Inordertoestimatethetemporalandspatialecologicalenvironmentimpactsofcascadehydropower,thethesisemploysthecascadehydropowerdevelopmentpilotproject,QingjiangRiverBasincascadehydropowerexploitation,asanexample.ThequantitativeremotesensinginversionmethodsandGIShavebeenadoptedtoqualitativelyandquantitativelyresearchtheimpactsofcascadehydropowerdevelopmentonsomekeyecologicalenvironmentfactors,suchaswaterenvironment,vegetationenvironmentandland-useenvironment.Summarily,themainstudiesandachievementsofthesisarelistedasfollowing:Firstly,theTM/ETM+remotelysensedimageriesprovidedbyseriesofsatellitesLandsathavebeenselectedasresearchoriginaldatainthisthesisbasedonanalyzingthecharacteristicsofmulti-sourceremotelysenseddataandtheexploitationofcascadehydropowerinQingjiangriverbasin.ThenthequantitativeremotesensinginversionIVtechniquesofkeyeco-environmentalfactors,thedatafusiontechniquesofmulti-sourcesremotelysenseddatabasedonremotesensing,themeansofspatialmanagementandspatialanalysisbasedonGIShavebeenresearchedinthisthesis.Secondly,thewaterextractionmodelsbasedonremotelysensedimageshavebeenanalyzedandtwowater-bodyextractionmodels,suchasnormalizeddifferencewaterindex(NDWI),modifiednormalizeddifferencewaterindex(MNDWI)havebeenemployedtoextractwaterresourceareafromtheselectedTM/EMT+imagessoastoquantitativelystudythespatiotemporalimpactofwaterdistributionenvironmentcausedbycascadehydropowerdevelopment.TheresultsshowthatthecascadehydropowerdevelopmenthassignificantinfluenceonthespatiotemporaldistributionofwaterresourcesinQingjiangriverbasin:beforeundevelopedin1987,thetotalwaterresourceareaofQingjiangmiddle-downstreamis65.31km2,andthatis102.25km2in2000,andthatreaches112.74km2in2004.Comparedwith1987,thetotalwaterresourceareain2004increases72.67%,however,thewaterresourcesincreaseregionsanddecreaseregionscoexistinQingjiangriverbasinduringthecourseofthecascadehydropowerdevelopment.Thirdly,thedimidiatepixelmodelbasedonnormalizeddifferencevegetationindex(NDVI)hasbeenadoptedtoestimatethevegetationfractionbasedonTM/ETM+imagessoastoquantitativelyresearchtheimpactofcascadehydropowerexploitationonthevegetationenvironment.TheresultsindicatethatthecascadehydropowerdevelopmenthassignificantimpactonthevegetationenvironmentinQingjiangriverbasin.ThevegetationcovertypesinQingjiangriverbasinincludethreemajortypes,suchasmiddlevegetationcoverage,middle-highvegetationcoverageandhighvegetationcoverage.In1987,theproportionofthreemajorvegetationcoveragetypesare23.93%,23.18%,25.43%respectively,andin2000,theproportionsofthreemajorvegetationcoveragetypesarefollowedby15.75%,21.44%and45.14%,whencometo2004,theproportionsofthreemajorvegetationcoveragetypesare11.66%,20.04%and55.84%respectively.Finally,supervised-classificationmethodbasedonmaximum-likelihoodandGeographicInformationSystem(GIS)havebeenusedtoquantitativelyresearchtheimpactofcascadehydropowerexploitationontheland-use/landcover-change(LUCC)environment.TheresultsilluminatethatthecascadehydropowerdevelopmenthasasignificantinfluenceontheLUCCenvironmentinQingjiangriverbasinfrom1987toV2004.From1987to2000,the1.77percentoffarmlandand29.24percentofurban-ruralresidentshavebeensubmergedbythecascadereservoirsaftercompletiontwocascadehydropowerstationsinQingjiangRiver.Fromthe2000to2004,the1.60percentoffarmlandand13.98percentofurban-ruralresidentshavebeensubmergedbythecascadereservoirs.Inaword,thecascadehydropowerdevelopmentpromotesregionaleconomicdevelopment,andsimultaneitythecascadehydropowerdevelopmentalsoacceleratestheconversio