|
OALib Journal期刊
ISSN: 2333-9721
费用：99美元
|
|
|
|
基于EVI指数的DMSP/OLS夜间灯光数据去饱和方法
DOI: 10.11821/dlxb201508012, PP. 1339-1350
Keywords: DMSP/OLS,夜间灯光,EANTLI,饱和,EVI指数
Abstract:
DMSP/OLS夜间灯光数据被广泛应用于表征人类活动强度及其生态环境影响的诸多研究中,但因OLS传感器设计的局限,在用电强度较高的城市中心,灯光信号存在明显的饱和,这一不足可能影响到一些基于夜间灯光数据研究成果的可靠性。针对这一问题,NOAA-NGDC研发了辐射定标算法,但因缺乏星上定标系统,算法较为复杂,且受较多条件限制等原因,目前只有部分时期的辐射定标数据产品（RCNTL）。近期有学者提出一种基于植被指数NDVI构建的城市灯光指数VANUI,为灯光数据去饱和研究提供了一个操作简单且结果良好的方法,但该方法在一些城市效果不佳。基于此,本文综合利用夜间灯光与EVI指数信息,通过对VANUI指数构建方法进行改进,建立了一个新的缓解夜间灯光强度饱和的EANTLI指数。为了评价指数的效果,将EANTLI与VANUI从三个方面进行比较①区分、识别饱和区内地物的能力;②与RCNTL的拟合程度;③对用电量估算的效果。结果表明EANTLI在三个方面均表现出优势,在潜在饱和区内对特征地物具有更高的可区分性,与RCNTL的线性相关程度更高,与用电量的相关性相比于NTL、VANUI亦明显提高。因此可以认为EANTLI在指数的设计上较为合理,不仅易于计算,而且能达到较好的缓解灯光强度饱和、凸现城市内部差异的目的,在用于反演城市发展指标时能获得更为准确的结果,因此具有较高的应用价值。
| [1] | LoCP.ModelingthepopulationofChinausingDMSPoperationallinescansystemnighttimedata.PhotogrammetricEngineeringandRemoteSensing,2001,67(9):1037-1047.Radiance-calibratedDMSP-OLSnighttimelightsdataofChinaacquiredbetweenMarch1996andJanuary-February1997wereevaluatedfortheirpotentialasasourceofpopulationdataattheprovincial,county,andcitylevels.Thelightclusterswereclassifiedintosixcategoriesoflightintensity,andtheirarea]extentswereextractedfromtheimage.Meanpixelvaluesoflightclusterscorrespondingtothesettlementswerealsoextracted.Alightvolumemeasurewasdevelopedtogaugethethree-dimensionalcapacityofasettlement.Adensityoflightclustermeasureknownaspercentlightareawasalsocalculatedforeachspatialunit.AllometricgrowthmodelsandlinearregressionmodelsweredevelopedtoestimatetheChinesepopulationandpopulationdensitiesatthethreespatiallevelsusinglightarea,lightvolume,pixelmean,andpercentlightareaasindependentvariables.ItwasfoundthattheDMSP-OLSnighttimedataproducedreasonablyaccurateestimatesofnon-agricultural(urban)populationatboththecountyandcitylevelsusingtheallometricgrowthmodelandthelightareaorlightvolumeasinput.Non-agriculturalpopulationdensitywasbestestimatedusingpercentlightareainalinearregressionmodelatthecountylevel.Thetotalsumsoftheestimatesfornon-agriculturalpopulationandevenpopulationoverallcloselyapproximatedthetruevaluesgivenbytheChinesestatisticsatallthreespatiallevels.Itisconcludedthatthe1-kmresolutionradiance-calibratedDMSP-OLSnighttimelightsimagehasthepotentialtoprovidepopulationestimatesofacountryandshedlightonitsurbanpopulationfromspace.DOI:10.1016/S0031-0182(01)00321-2Crossref
|
| [2] | LoCP.UrbanindicatorsofChinafromradiance-calibrateddigitalDMSP-OLSnighttimeimages.AnnalsoftheAssociationofAmericanGeographers,2002,92(2):225-240.Acompositeofcloud-freeradiance-calibratedDefenseMeteorologicalSatelliteProgram(DMSP)-OperationalLinescanSystem(OLS)nighttimeimagesofChinaacquiredbetweenMarch1996andJanuary-February1997in30-arc-secondgridsinbyteformatisevaluatedforitsusefulnessinextractingurbanindicatorsdataofChina.Withtheaidofimage-processingandEnvironmentalSystemsResearchInstitute'sArcViewGISsoftwarewith3-DAnalysisextension,zonalvariationsofradiancewereextractedandthree-dimensionalmodelscreatedusingTriangulatedIrregularNetwork(TIN)functionalityforthirty-fivecities.TwovariableswereextractedfromtheTINmodel:surfaceareaandvolume.ThesewereusedseparatelyasindependentvariablesintheformofanallometricgrowthmodeltoestimatethefollowingurbanindicatorsforindividualChinesecities:nonagriculturalpopulation,grossdomesticproduct(GDP),built-uparea,andelectricityconsumption.TheestimatesobtainedwerecheckedagainstdatasuppliedbytheStatisticalBureauofChinafor1997,thesameperiodduringwhichtheDMSP-OLSimageswereacquired.Itwasfoundthaturbanindicatorsofacceptableaccuracycouldbeobtained,providedthatcitiesdesignatedasSpecialEconomicZones(SEZs)wereexcluded.Volumeappearedtobeslightlybetterthansurfaceareainestimatingtheseurbanindicators.Inallcases,totalnonagriculturalpopulationforallthecitiescombinedcouldbeveryaccuratelyestimated,indicatingtheusefulnessoftheradiance-calibratedDMSP-OLSnighttimeimagestodeterminethelevelofurbanizationinChina.Thethree-dimensionalTINmodelportraysanilluminatedurbanarea(IUA)dome,theshapeofwhichisaffectedbytheinternalstructureofthecity.ThecoreofthecityinvariablyexhibitsthehighestleveloflightintensityintheDMSP-OLSnighttimeimage.However,onlyaverylimitedamountofland-useinformationcouldbeextracted.TheimagewasalsofoundtobeusefulindetectingthespecialurbanphenomenoninChinaknownastheextendedmetropolitanregion,inwhichruralandurbanusesintermixedandsmalltownsdeveloped.Inconclusion,despitethelowspatialresolution,theradiance-calibratedDMSP-OLSnighttimeimagesarecapableofprovidingusefuldemographicandsocioeconomicdataforcities.DOI:10.1111/1467-8306.00288Wiley
|
| [3] | SuttonPC,RobertsD,ElvidgeCD,etal.Censusfromheaven:Anestimateoftheglobalhumanpopulationusingnight-timesatelliteimagery.InternationalJournalofRemoteSensing,2001,22(16):3061-3076.Night-timesatelliteimageryprovidedbytheDefenseMeteorologicalSatelliteProgram'sOperationalLinescanSystem(DMSPOLS)isevaluatedasameansofestimatingthepopulationofallthecitiesoftheworldbasedontheirarealextentintheimage.Aglobalnight-timeimageproductwasregisteredtoadatasetof2000knowncitylocationswithknownpopulations.ArelationshipbetweenarealextentandcitypopulationdiscoveredbyToblerandNordbeckisidentifiedonanationbynationbasistoestimatethepopulationofthe22920urbanclustersthatexistinthenight-timesatelliteimage.Therelationshipbetweencitypopulationandcityarealextentwasderivedfrom1597citypointlocationswithknownpopulationthatlandedina'lit'areaoftheimage.Duetoconurbation,these1597citiesresultedinonly1383pointsofanalysisforperformingregression.Whenseveralcitiesfellintoone'lit'areatheirpopulationsweresummed.Theresultsofthisanalysisallowforanestimateoftheurbanpopulationofeverynationoftheworld.Byusingtheknownpercentofpopulationinurbanareasforeverynationatotalnationalpopulationwasalsoestimated.Thesumoftheseestimatesisatotalestimateoftheglobalhumanpopulation,whichinthiscasewas6.3billion.Thisisfairlyclosetothegenerallyacceptedcontemporaneous(1997)estimateoftheglobalpopulationwhichstoodatapproximately5.9billion.DOI:10.1080/01431160010007015Taylor&Francis
|
| [4] | SuttonPC.Ascale-adjustedmeasureof"Urbansprawl"usingnighttimesatelliteimagery.RemoteSensingofEnvironment,2003,86(3):353-369.“UrbanSprawl”isagrowingconcernofcitizens,environmentalorganizations,andgovernments.Negativeimpactsoftenattributedtourbansprawlaretrafficcongestion,lossofopenspace,andincreasedpollutantrunoffintonaturalwaterways.Definitionsof“UrbanSprawl”rangefromlocalpatternsoflanduseanddevelopmenttoaggregatemeasuresofpercapitalandconsumptionforgivencontiguousurbanareas(UA).ThisresearchcreatesameasureofpercapitalanduseconsumptionasanaggregateindexforthespatiallycontiguousurbanareasoftheconterminousUnitedStateswithpopulationof50,000orgreater.NighttimesatelliteimageryobtainedbytheDefenseMeteorologicalSatelliteProgram'sOperationalLinescanSystem(DMSPOLS)isusedasaproxymeasureofurbanextent.Thecorrespondingpopulationoftheseurbanareasisderivedfromagridoftheblockgroupleveldatafromthe1990U.S.Census.ThesenumbersareusedtodeveloparegressionequationbetweenLn(UrbanArea)andLn(UrbanPopulation).The‘scale-adjustment’mentionedinthetitlecharacterizesthe“UrbanSprawl”ofeachoftheurbanareasbyhowfaraboveorbelowtheyareonthe“SprawlLine”determinedbythisregression.This“SprawlLine”allowsforamorefaircomparisonof“UrbanSprawl”betweenlargerandsmallermetropolitanareasbecauseasimplemeasureofpercapitalandconsumptionorpopulationdensitydoesnotaccountforthenaturalincreaseinaggregatepopulationdensitythatoccursascitiesgrowinpopulation.Citiesthathavemore“UrbanSprawl”bythismeasuretendedtobeinlandandMidwesterncitiessuchasMinneapolis–St.Paul,Atlanta,Dallas–Ft.Worth,St.Louis,andKansasCity.Surprisingly,westcoastcitiesincludingLosAngeleshadsomeofthelowestlevelsof“UrbanSprawl”bythismeasure.Thereweremanylowlightlevelsseeninthenighttimeimageryaroundthesemajorurbanareasthatwerenotincludedineitherofthetwodefinitionsofurbanextentusedinthisstudy.Theseareasmayrepresentagrowingcommuter-shedofurbanworkerswhodonotliveintheurbancorebutnonethelesscontributetomanyoftheimpactstypicallyattributedto“UrbanSprawl”.“UrbanSprawl”isdifficulttodefinepreciselypartlybecausepublicperceptionofsprawlislikelyderivedfromlocallanduseplanningdecisions,spatio-demographicchangeingrowingurbanareas,andchangingvaluesandsocialmoresresultingfromdifferentialratesofinternationalmigrationtotheurbanareasoftheUnitedStates.Nonetheless,theaggregatemeasuresderivedherearesomewhatdifferentthansimilarpreviouslyusedmeasuresinthattheyare‘scale-adjusted’;also,thespatialpatternsof“UrbanSprawl”shownhereshedsomeinsightandraiseinterestingquestionsabouthowthedynamicsof“UrbanSprawl”arechanging.Elsevier
|
| [5] | ZhuoLi,ChenJin,ShiPeijun,etal.ModelingpopulationdensityofChinain1998basedonDMSP/OLSnighttimelightimage.ActaGeographicaSinica,2005,60(2):266-276.Spatialdistributionofpopulationdensityiscrucialforanalyzingtherelationshipamongeconomicgrowth,environmentprotectionandresourceutilization.Inthisstudy,populationdensityofChinain1998at1-kmresolutiongridswassimulatedbyintegratingDMSP/OLSnon-radiancecalibratednighttimelightimage,SPOT/VEGETATION10-daymaximumNDVIdata,populationcensusdataandvectordataofcountyboundary.Notonlythepopulationdensityinlightpatchesbutalsothatoutofthemwasestimatedinfourtypesofareas.Foreacharea,inlightpatches,themodelforpopulationdensityestimationwasdevelopedbasedonthesignificantcorrelationbetweenlightintensityan原文[卓莉,陈晋,史培军,等.基于夜间灯光数据的中国人口密度模拟.地理学报,2005,60(2):266-276.]原文
|
| [6] | ElvidgeCD,BaughKE,KihnEA,etal.Relationbetweensatelliteobservedvisible-nearinfraredemissions,population,economicactivityandelectricpowerconsumption.InternationalJournalofRemoteSensing,1997,18(6):1373-1379.Thearealitbyanthropogenicvisible-nearinfraredemissions(i.e.,lights)hasbeenestimatedfor21countriesusingnight-timedatafromtheDefenseMeteorologicalSatelliteProgram(DMSP)OperationalLinescanSystem(OLS).Thearealitishighlycorrelatedtogrossdomesticproductandelectricpowerconsumption.Significantoutliersexistintherelationbetweenarealitandpopulation.TheresultsindicatethatthelocallevelofeconomicdevelopmentmustbefactoredintotheapportionmentofpopulationacrossthelandsurfacebasedonDMSP-OLSobservedlights.DOI:10.1080/014311697218485Taylor&Francis
|
| [7] | AmaralS,CamaraG,MonteiroAMV,etal.EstimatingpopulationandenergyconsumptioninBrazilianAmazoniausingDMSPnight-timesatellitedata.Computers,EnvironmentandUrbanSystems,2005,29(2):179-195.ThispaperdescribesamethodologytoassesstheevidenceofhumanpresenceandhumanactivitiesintheBrazilianAmazoniaregionusingDMSP/OLSnight-timesatellitesensorimagery.ItconsistsonexploringthepotentialofthesensordataforregionalstudiesanalysingthecorrelationbetweenDMSPnight-timelightfociandpopulation,andthecorrelationbetweenDMSPnight-timelightfociandelectricalpowerconsumption.InthemosaicofDMSP/OLSnight-timelightimageryfromSeptember1999,248townsweredetectedfromatotalof749muniDOI:10.1016/j.compenvurbsys.2003.09.004Elsevier
|
| [8] | WelchR.Monitoringurbanpopulationandenergyutilizationpatternsfromsatellitedata.RemoteSensingofEnvironment,1980,9(1):1-9.UrbanpopulationtrendsinChinaandenergyutilizationpatternsintheUnitedStateshavebeenassessedfromLANDSATandDefenseMeteorologicalSatelliteProgram(DMSP)images.Regressionmodelsdevelopedfrompopulation(P)dataandurbanarea(A)measurementsonLANDSATimagesareoftheformP=$\alphaA^b$andprovideinsightsintothesuccessofChineseurbanplanningpoliciesforcitieswithpopulationsof500,000to2,000,000people.Studiesoftherelationshipsbetweenpopulation,urbanarea,andelectric-energyutilizationpatternshavebeenconductedfromDMSPimagesoftheUnitedStates.Microdensitometerprofilesofilluminatedcitiesrecordedonnighttime(visualband)imagesareusedincombinationwiththemapboundariesofthebuilt-upareastocreateuniquethree-dimensionalrepresentationsoftheurbancenters.Thevolumesofthesethree-dimensionalfiguresmaybecomputedandplottedwithrespecttopopulationand/orenergyutilizationdatatomodelregionalpatternsofuse.Improvementsinthequalityofsensordatashouldincreasethepotentialformonitoringurbanenergydemandsandpopulations.DOI:10.1016/0034-4257(80)90043-7Elsevier
|
| [9] | ZhuoLi,IchinoseT,ZhengJing,etal.ModelingthepopulationdensityofChinaatthepixellevelbasedonDMSP/OLSnon-radiance-calibratednight-timelightimages.InternationalJournalofRemoteSensing,2009,30(4):1003-1018.
|
| [10] | RaupachMR,RaynerPJ,PagetM.Regionalvariationsinspatialstructureofnightlights,populationdensityandfossil-fuelCO2emissions.EnergyPolicy,2010,38(9):4756-4764.Elsevier
|
| [11] | SuttonPC,ElvidgeCD,GhoshT.Estimationofgrossdomesticproductatsub-nationalscalesusingnighttimesatelliteimagery.InternationalJournalofEcologicalEconomics&Statistics,2007,8(7):5-21.Thereisadegreeofuncertaintyinthemeasurementand/orvalidationofnationalandsub-nationaleconomicdatasuchasGrossDomesticProduct(GDP).InsomecircumstancesitcanbeveryusefultohavealternativemeasuresofnumberslikeGDPtoprovideevidenceforthevalidationorinvalidationofclaimsofsomenationsorregionsregardingtheireconomicproductivity.ThisresearchexploresthefeasibilityofdevelopingpredictiverelationshipsbetweenobservedchangesinnighttimesatelliteimagesderivedfromtheDefenseMeteorologicalSatelliteProgramOperationalLinescanSystem(DMSP-OLS),andchangesinpopulationandGrossDomesticProduct(GDP).PDF
|
| [12] | GhoshT,PowellRL,ElvidgeCD,etal.Sheddinglightontheglobaldistributionofeconomicactivity.TheOpenGeographyJournal,2010,3(1):148-161.Collectionofdataoneconomicvariables,especiallysub-nationalincomelevels,isproblematic,duetovariousshortcomingsinthedatacollectionprocess.Additionally,theinformalecoPDFCrossref
|
| [13] | GhoshT,AndersonS,PowellR,etal.EstimationofMexico'sinformaleconomyandremittancesusingnighttimeimagery.RemoteSensing,2009,1(3):418-444.ALib
|
| [14] | HanXiangdi,ZhouYi,WangShixin,etal.GDPspatializationinChinabasedonDMSP/OLSdataandlandusedata.RemoteSensingTechnologyandApplication,2012,27(3):396-405.GDPisakeyindicatorofsocioeconomicdevelopment,urbanplanning,andenvironmentalprotection,accurateestimatesofthemagnitudeandspatialdistributionofeconomicactivityhavemanyusefulapplicationsinresourcesandenvironmentalsciences.Developingalternativemethodsmayprovetobeusefulformakingestimatesofgrossdomesticproductwhenothermeasuresareofsuspectaccuracyorunavailable.Basedonthesummaryandanalysisofexistingeconomicactivityspatializationapproaches,thispaperexploredthepotentialforspatializingGDPthroughChinausingnight-timesatelliteimagery(DMSP/OLS)andland-usedata.IncreatingtheGDPlinearregressionmodelofsecondaryindustryandtertiaryindustry,night-timelightintensityandlitareas,underdifferenttypesoflanduse,wereemployedaspredictorvariables,andtheGDPstatisticaldatawasasdependentvariable,meanwhile,modelofprimaryindustrybasedonthelandusedata.Toimprovemodelperformance,31zoneswerecreatedaccordingtoprovincialadministrativeboundary.Themodelofprimaryindustryisobservedtohaveacorrelation(R2)rangingfrom0.7to0.95inmajorityzonesandR2ofsecondaryindustryandtertiaryindustrymodleisrangingfrom0.8to0.98inmajorityzones.AcomparisonoftheresultsofthisresearchwithotherresearchesshowsthatspatializedGDPdensitymap,preparedonnight-timeimageryandland-usedata,whichreflectstheGDPdistributioncharacteristicsmoreexplicitlyandgreaterdetail.Meantime,thedensitymapissignificantsustainableeconomicdevelopmentpoliciesandbasicallyexplorestherelationshipbetweensocioeconomicandregionalecologicalenvironmentinteraction.原文[韩向娣,周艺,王世新,等.基于夜间灯光和土地利用数据的GDP空间化.遥感技术与应用,2012,27(3):396-405.]GDP是社会经济发展、区域规划和资源环境保护的重要指标之一。然而,传统以各级行政单元为主的GDP统计资料无法显示区域内部GDP的差异,尝试通过GDP空间化来解决这个问题,以满足资源环境研究领域对空间型社会经济数据的需求。在分析总结国内外社会经济数据空间化技术方法的基础上,第一产业基于土地利用数据建模、第二产业和第三产业基于DMSP/OLS与土地利用数据结合生成的土地灯光参数建模。为提高模型质量,全国按照省级行政边界分区,将因变量GDP分产业分区建模,第一产业绝大部分区域的模型精度在0.7～0.95之间,第二产业和第三产业绝大部分区域的模型精度在0.8～0.98之间。通过与其他GDP空间化技术和结果的比较分析,本研究中的GDP空间化方法无论是模型精度还是GDP密度分布结果都具有一定的优势。生成的GDP密度图能较完整地反映全国GDP分布细节以及宏观分布特征,可为将来经济策略和发展路线的绘制提供一定依据。原文
|
| [15] | ElvidgeCD,BaughKE,AndersonSJ,etal.TheNightLightDevelopmentIndex(NLDI):Aspatiallyexplicitmeasureofhumandevelopmentfromsatellitedata.SocialGeography,2012,7(1):23-35.Wehavedevelopedasatellitedataderived''NightLightDevelopmentIndex''(NLDI)asasimple,objective,spatiallyexplicitandgloballyavailableempiricalmeasurementofhumandevelopmentderivedsolelyfromnighttimesatelliteimageryandpopulationdensity.Thereisincreasingrecognitionthatthedistributionofwealthandincomeamongstthepopulationinanationorregioncorrelatesstronglywithboththeoverallhappinessofthatpopulationandtheenvironmentalqualityofthatnationorregion.Measuringthedistributionofwealthandincomeatnationalandregionalscalesisaninterestingandchallengingproblem.GinicoefficientsderivedfromLorenzcurvesareawell-establishedmethodofmeasuringincomedistribution.Nonetheless,therearemanyshortcomingsoftheGinicoefficientasameasureofincomeorwealthdistribution.Ginicoefficientsaretypicallycalculatedusingnationalleveldataonthedistributionofincomethroughthepopulation.Suchdataarenotavailableformanycountriesandtheresultsaregenerallylimitedtosinglevaluesrepresentingentirecountries.Inthispaperwedevelopanindexfortheco-distributionofnocturnallightandpeoplethatisderivedwithouttheuseofmonetarymeasuresofwealthandiscapableofprovidingaspatialdepictionofdifferencesindevelopmentwithincountries.DOI:10.5194/sg-7-23-2012ALib
|
| [16] | ChenXi,NordhausWD.Usingluminositydataasaproxyforeconomicstatistics.ProceedingsoftheNationalAcademyofSciences,2011,108(21):8589-8594.Apervasiveissueinsocialandenvironmentalresearchhasbeenhowtoimprovethequalityofsocioeconomicdataindevelopingcountries.Giventheshortcomingsofstandardsources,thepresentstudyexaminesluminosity(measuresofnighttimelightsvisiblefromspace)asaproxyforstandardmeasuresofoutput(grossdomesticproduct).Wecompareoutputandluminosityatthecountrylevelandatthe1°latitude×1°longitudegrid-celllevelfortheperiod1992–2008.Wefindthatluminosityhasinformationalvalueforcountrieswithlow-qualitystatisticalsystems,particularlyforthosecountrieswithnorecentpopulationoreconomiccensuses.PNAS
|
| [17] | WangQi,YuanTao,ZhengXinqi.GDPgrossanalysisatprovince-levelinChinabasedonnighttimelightsatelliteimagery.UrbanDevelopmentStudies,2013,20(7):44-48.CNKI[王琪,袁涛,郑新奇.基于夜间灯光数据的中国省域GDP总量分析.城市发展研究,2013,20(7):44-48.]国内生产总值（GDP）通常以各种统计资料、统计报表为基础进行核算，费时耗财。本文提出了一种基于DMSP／OLS夜间灯光数据测算GDP的新方法。在分析夜间灯光强度与GDP关系基础上，采用回归分析的方法建立了GDP空间化模型，在GIS平台中实现了省级GDP空间化模拟。结果显示，建立的回归模型拟合精度达到0．964，显著性水平低于0．05，通过夜间灯光数据计量省级GDP总量的效率币口精度都令人满意，将灯光数据应用于经济分析具有可行性。进一步修正完善模型，将为区域经济研究提供一种新研究方法。DOI:10.3969/j.issn.1006-3862.2013.07.007CNKI
|
| [18] | ChandTRKiran,BadarinathKVS,ElvidgeCD,etal.SpatialcharacterizationofelectricalpowerconsumptionpatternsoverIndiausingtemporalDMSP-OLSnight-timesatellitedata.InternationalJournalofRemoteSensing,2009,30(3):647-661.Changesinelectricpowerconsumptionpatternsofacountryoveraperiodoftimereflectonitssocio-economicdevelopmentandenergyutilizationprocesses.Inthepresentstudy,wecharacterizedspatialandtemporalchangesinelectricpowerconsumptionpatternsoverIndiaduring1993to2002,using'night-timelights'datagivenbytheDefenseMeteorologicalSatelliteProgram-OperationalLineScanSystem(DMSP-OLS)overtheIndianregion.TheOLSoperatesintwobands:visible(0.5-0.90208m)andthermal(10.5-12.50208m)andhasauniquecapabilityofpickingupfaintsourcesofvisibDOI:10.1080/01431160802345685Taylor&Francis
|
| [19] | ChalkiasC,PetrakisM,PsiloglouB,etal.ModellingoflightpollutioninsuburbanareasusingremotelysensedimageryandGIS.JournalofEnvironmentalManagement,2006,79(1):57-63.Thispaperdescribesamethodologyformodellinglightpollutionusinggeographicalinformationsystems(GIS)andremotesensing(RS)technology.Theproposedapproachattemptstoaddresstheissueofenvironmentalassessmentinsensitivesuburbanareas.Themodernwayoflifeindevelopingcountriesisconductivetoenvironmentaldegradationinurbanandsuburbanareas.Onespecificparameterforthisdegradationislightpollutionduetointenseartificialnightlighting.ThispaperaimstoassessthisparameterfortheAthensmetropolitanarea,usingmodernanalyticalanddatacapturingtechnologies.ForthisDOI:10.1016/j.jenvman.2005.05.015PMID:16171928Elsevier
|
| [20] | HeChunyang,MaQun,LiTong,etal.SpatiotemporaldynamicsofelectricpowerconsumptioninChineseMainlandfrom1995to2008modeledusingDMSP/OLSstablenighttimelightsdata.JournalofGeographicalSciences,2012,22(1):125-136.Electricpowerconsumption(EPC)isoneofthebasicindicesforevaluatingelectricpoweruse.ObtainingtimelyandaccuratedataonthespatiotemporaldynamicsofEPCiscrucialforunderstandingandpracticaldeploymentofelectricpowerresources.Inthisstudy,anEPCmodelwasdevelopedusingstablenighttimelightstime-seriesdatafromtheDefenseMeteorologicalSatelliteProgramOperationalLinescanSystem(DMSP/OLS).ThemodelwasusedtoreconstructthespatialpatternsofEPCinChineseMainlandatthecountylevelfrom199DOI:10.1007/s11442-012-0916-3原文
|
| [21] | DollCH,MullerJ,ElvidgeCD.Night-timeimageryasatoolforglobalmappingofsocioeconomicparametersandgreenhousegasemissions.AMBIO:aJournaloftheHumanEnvironment,2000,29(3):157-162.Night-timesatelliteimageryacquiredbetweenOctober1994andMarch1995ishereexploitedtoderiveeconomicandenergy-relatedglobalmaps.Byconsideringthelitareaofacity,andcombiningthiswithancillarystatisticalinformation,ananalysiswasperformeddesignedtoinvestigatethepotentialofnight-timeimageryforquantitativeestimationofglobalsocioeconomicparameters.Anattempttoestimateglobalurbanpopulationusingcorrelationsoflitareaandurbanpopulationstoderivecountry-levelrelationshipsaccountedforover90%ofthequotedtotal.Furthermore,thetotallitareaofacountryhasastatisticallysignificanthighcorrelationvaluewithotherparameters,specificallyGrossDomesticProduct(GDP)andtotalcarbondioxide(CO2)emission.Thenewtechniqueislimitedbythespatialresolutionofthesensorandthepoorercorrelationusingnight-timelightsforcentrally-plannedeconomies.Thesefindingsoffergreatpotentialforsynopticglobalmappingofsuchparametersinthefuture.Ambio,Vol.29,No.3,pp.157-162,2000DOI:10.1639/0044-7447(2000)029[0157:NTIAAT]2.0.CO;2CrossrefPDF
|
| [22] | LiJinggang,HeChunyang,ShiPeijun,etal.Theuseofmultisourcesatelliteandgeospatialdatatostudytheecologicaleffectsofurbanization:AcaseoftheurbanagglomerationsinBohaiRim.JournalofRemoteSensing,2007,11(1):115-126.[李景刚,何春阳,史培军,等.基于DMSP/OLS灯光数据的快速城市化过程的生态效应评价研究:以环渤海城市群地区为例.遥感学报,2007,11(1):115-126.]维普
|
| [23] | ImhoffML,TuckerL,ComptonJ,etal.TheuseofmultisourcesatelliteandgeospatialdatatostudytheeffectofurbanizationonprimaryproductivityintheUnitedStates.IEEETransactionsonGeoscienceandRemoteSensing,2000,38(6):2549-2556.
|
| [24] | ImhoffML,BounouaL,DeFriesR,etal.TheconsequencesofurbanlandtransformationonnetprimaryproductivityintheUnitedStates.RemoteSensingofEnvironment,2004,89(4):434-443.WeusedatafromtwosatellitesandaterrestrialcarbonmodeltoquantifytheimpactofurbanizationonthecarboncycleandfoodproductionintheUSasaresultofreducednetprimaryproductivity(NPP).OurresultsshowthaturbanizationistakingplaceonthemostfertilelandsandhencehasadisproportionatelylargeoverallnegativeimpactonNPP.UrbanlandtransformationintheUShasreducedtheamountofcarbonfixedthroughphotosynthesisby0.04pgperyearor1.6%ofthepre-urbaninput.Thereductionisenoughtooffsetthe1.8%gainmadebytheconversionoflandtoagriculturaluse,eventhoughurbanizationcoversanarealessthan3%ofthelandsurfaceintheUSandagriculturallandsapproach29%ofthetotallandarea.Atlocalandregionalscales,urbanizationincreasesNPPinresource-limitedregionsandthroughlocalizedwarming“urbanheat”contributestotheextensionofthegrowingseasonincoldregions.IntermsofbiologicallyavailDOI:10.1016/j.rse.2003.10.015Elsevier
|
| [25] | ElvidgeCD,BaughK,DietzJB,etal.RadiancecalibrationofDMSP-OLSlow-lightimagingdataofhumansettlements.RemoteSensingofEnvironment,1999,68(1):77-88.Nocturnallightingisaprimarymethodforenablinghumanactivity.Outdoorlightingisusedextensivelyworldwideinresidential,commercial,industrial,publicfacilities,androadways.AradiancecalibratednighttimelightsimageoftheUnitedStateshasbeenassembledfromDefenseMeteorologicalSatelliteProgram(DMSP)OperationalLinescanSystem(OLS).Thesatelliteobservationofthelocationandintensityofnocturnallightingprovideauniqueviewofhumanitiespresenceandcanbeusedasaspatialindicatorforothervariablesthataremoredifficulttoobserveataglobalscale.Examplesincludethemodelingofpopulationdensityandenergyrelatedgreenhousegasemissions.DOI:10.1016/S0034-4257(98)00098-4Elsevier
|
| [26] | ZiskinD,BaughK,HsuFC.Methodsusedforthe2006radiancelights.ProceedingsoftheAsiaPacificAdvancedNetwork,2010:131-142.TheOperationalLinescanSystem(OLS)flownontheDefenseMeteorologicalSatelliteProgram(DMSP)satellites,hasauniquecapabilitytorecordlowlightimagingdataatnightworldwide.ThesedataarearchivedattheNationalOceanicandAtmosphericAdministration(NOAA)NationalGeophysicalDataCenter(NGDC).Theusefuldatarecordstretchesbackto1992andisongoing.TheOLSvisiblebanddetectorobservesradiancesaboutonemilliontimesdimmerthanmostotherEarthobservingsatellites.Thesensoristypicallyoperatedinahighgainsettingtoenablethedetectionofmoonlitclouds.However,withsixbitquantizationandlimiteddynamicrange,therecordeddataaresaturatedinthebrightcoresofurbancenters.Alimitedsetofobservationshavebeenobtainedatlowlunarilluminationwereobtainedwherethegainofthedetectorwassetsignificantlylowerthanitstypicaloperationalsetting(sometimesbyafactorof100).Bycombiningthesesparsedataacquiredatlowgainsettingswiththeoperationaldataacquiredathighgainsettings,wehaveproducedaglobalnighttimelightsproductfor2006withnosensorsaturation.Thisproductcanberelatedtoradiancesbasedonthepre-flightssensorcalibration.DOI:10.7125/APAN.30.18Crossref
|
| [27] | LetuH,HaraM,YagiH,etal.Estimatingenergyconsumptionfromnight-timeDMPS/OLSimageryaftercorrectingforsaturationeffects.InternationalJournalofRemoteSensing,2010,31(16):4443-4458.Amethodologyispresentedtoaccuratelyestimateelectricpowerconsumptionfromsaturatednight-timeDefenseMeteorologicalSatelliteProgram(DMSP)OperationalLinescanSystem(OLS)imageryusingastablelightcorrection.AnareacorrectionforthestablelightimageofDMSP/OLSfortheyear1999wasperformedandthebuild-uparearatedatawereusedtoclarifytheintensitydistributioncharacteristicsofthestablelight.Basedonthespatialdistributioncharacteristicsofthestablelight,thesaturationlightoftheelectricpowersupplyareaofJapanwascorrectedusingacubicregressionequation.TheregressionbetweenthecorrectioncalculationsbythecubicregressionequationandthestatisticalelectricpowerconsumptiondatawasappliedinJapanandalsoinChina,Indiaand10otherAsiancountries.Thecorrectionmethodwasthenevaluated.Thisstudyconfirmsthatelectricpowerconsumptioncanbeestimatedwithhighprecisionfromthestablelight.DOI:10.1080/01431160903277464Taylor&Francis
|
| [28] | LetuH,HaraM,TanaG,etal.AsaturatedlightcorrectionmethodforDMSP/OLSnighttimesatelliteimagery.IEEETransactionsonGeoscienceandRemoteSensing,2012,50(2):389-396.SeveralstudieshaveclarifiedthatelectricpowerconsumptioncanbeestimatedfromtheDefenseMeteorologicalSatelliteProgram/OperationalLinescanSystem(DMSP/OLS)stablelightimagery.Asdigitalnumbers(DNs)ofstablelightimagesareoftensaturatedinthecenterofcityareas,wedevelopedasaturatedlightcorrectionmethodfortheDMSP/OLSstablelightimageusingthenighttimeradiancecalibrationimageoftheDMSP/OLS.Thecomparisonbetweenthenonsaturatedpartofthestablelightimagefor1999andtheradiancecalibrationimagefor1996-1997inmajorareasofJapanshowedastronglinearcorrelation(R=92.73)betweentheDNsofbothimages.SaturatedDNsofthestablelightimagecouldthereforebecorrectedbasedonthecorrelationequationbetweenthetwoimages.Toevaluatethenewsaturatedlightcorrectionmethod,aregressionanalysisisperformedbetweenstatisticdataofelectricpowerconsumptionfromlightingandthecumulativeDNsofthestablelightimagebeforeandaftercorrectingforthesaturationeffectsbythenewmethod,incomparisontotheconventionalmethod,whichis,thecubicregressionequationmethod.Theresultsshowastrongerimprovementinthedeterminationcoefficientwiththenewsaturatedlightcorrectionmethod(R=0.91,P=1.7·10<;0.05)thanwiththeconventionalmethod(R=0.81,P=2.6·10<;0.05)fromtheinitialcorrelationwiththeuncorrecteddata(R=0.70,P=4.5·10<;0.05).Thenewmethodprovesthereforetobeveryefficientforsaturatedlightcorrection.DOI:10.1109/TGRS.2011.2178031EEEXplorePDF
|
| [29] | ZhangQingling,SchaafC,SetoKC.TheVegetationadjustedNTLUrbanIndex:Anewapproachtoreducesaturationandincreasevariationinnighttimeluminosity.RemoteSensingofEnvironment,2013,129:32-41.Thescienceandpolicycommunitiesincreasinglyrequireinformationaboutinter-urbanvariabilityinform,infrastructure,andenergyuseforcitiesgloballyandinatimelymanner.Nighttimelight(NTL)datafromtheDefenseMeteorologicalSatelliteProgram/OperationalLinescanSystem(DMSP/OLS)areabletoprovideinformationonnighttimeluminosity,acorrelateofthebuiltenvironmentandenergyconsumption.AlthoughNTLdataareusedtomapaggregatemeasuresofurbanareassuchastotalareaextent,theirabilitytocharacterizeinter-urbanvariationislimitedduetosaturationofthedatavalues,especiallyinurbancores.Hereweproposeanewspectralindex,theVegetationAdjustedNTLUrbanIndex(VANUI),whichcombinesMODISNDVIwithNTL,toachievethreekeygoals.First,theindexreducestheeffectsofNTLsaturation.Second,theindexincreasesvariationoftheNTLsignal,especiallywithinurbanareas.Third,theindexcorrespondstobiophysicalandurbancharacteristics.Additionally,theindexisintuitive,simpletoimplement,andenablesrapidcharacterizationofinter-urbanvariabilityinnighttimeluminosity.AssessmentsofVANUIshowthatitsignificantlyreducesNTLsaturationandincreasesvariationofdatavaluesincoreurbanareas.Assuch,VANUIcanbeusefulforstudiesofurbanstructure,energyuDOI:10.1016/j.rse.2012.10.022Elsevier
|
| [30] | PozziF,SmallC.AnalysisofurbanlandcoverandpopulationdensityintheUnitedStates.PhotogrammetricEngineeringandRemoteSensing,2005,71(6):719-726.InthisstudyweinvestigatethequestionofwhetherurbanandsuburbanareasintheUnitedStatescanbedefinedonthebasisofdemographicand/orphysicalcharacteristics,inparticular,populationdensityandvegetationabundance.WeinvestigatetheirrelationshipinthecitiesofAtlanta,Chicago,LosAngeles,NewYork,Phoenix,andSeattleandcomparetheresultswiththeUSGSNationalLandCoverDataset'surbanclasses.ThebimodaldistributionofU.S.populationdensityprovidesademographicbasisfordistinguishingruralandsuburbanlanduse,whileadistincttailofhighpopulationdensities(>10,000people/km(2))correspondstohighintensityurbanresidentialcores.Resultsshowthatthemaximumvegetationfractiondiminisheswithincreasingpopulationdensity,butthespectralheterogeneityatpixelscalesstillresultsinawiderangeofvegetationfractionswithindemographicallyurbanandsuburbanareas.NoneoftheUSGSresidentialclassesshowastrongcorrespondencetoeithervegetationfractionorpopulationdensity.However,quantitativecharacterizationofvegetationabundanceprovidesabasisforcomparisonofthephysicalenvironmentsOfsuburbanareas.Wesuggestthatclassificationschemesbasedonspectralheterogeneityatmultiplepixelscales,supplementedbyauxiliarydatasources,mayprovideamoreaccurateandself-consistentwaytoquantifyurbanlanduseandanalyzeurbangrowththantraditionalthematicclassificationschemes.DOI:10.1007/s11069-004-6485-8CrossrefPDF
|
| [31] | SmallC.Estimationofurbanvegetationabundancebyspectralmixtureanalysis.InternationalJournalofRemoteSensing,2001,22(7):1305-1334.Thespatio-temporaldistributionofvegetationisafundamentalcomponentoftheurbanenvironmentthatcanbequantifiedusingmultispectralimagery.However,spectralheterogeneityatscalescomparabletosensorresolutionlimitstheutilityofconventionalhardclassificationmethodswithmultispectralreflectancedatainurbanareas.Spectralmixturemodelsmayprovideaphysicallybasedsolutiontotheproblemofspectralheterogeneity.TheobjectiveofthisstudyistoexaminetheapplicabilityoflinearspectralmixturemodelstotheestimationofurbanvegetationabundanceusingLandsatThematicMapper(TM)data.TheinherentdimensionalityofTMimageryoftheNewYorkCityareasuggeststhaturbanreflectancemeasurementsmaybedescribedbylinearmixingbetweenhighalbedo,lowalbedoandvegetativeendmembers.Athree-componentlinearmixingmodelprovidesstable,consistentestimatesofvegetationfractionforbothconstrainedandunconstrainedinversionsofthreedifferentendmemberensembles.Quantitativevalidationusingvegetationabundancemeasurementsderivedfromhigh-resolution(2m)aerialphotographyshowsagreementtowithinfractionalabundancesof0.1forvegetationfractionsgreaterthan0.2.IncontrasttotheNormalisedDifferenceVegetationIndex(NDVI),vegetationfractionestimatesprovideaphysicallybasedmeasureofarealvegetationabundancethatmaybemoreeasilytranslatedtoconstraintsonphysicalquantitiessuchasvegetativebiomassandevapotranspiration.DOI:10.1080/01431160151144369Taylor&Francis
|
| [32] | WengQihao,LuDengsheng,LiangBingqing.Urbansurfacebiophysicaldescriptorsandlandsurfacetemperaturevariations.PhotogrammetricEngineeringandRemoteSensing,2006,72(11):1275-1286.Inremotesensingstudiesoflandsurfacetemperatures(LST),thematicland-useandland-cover(LULC)dataarefrequently-employedforsimplecorrelationanalysesbetweenLULCtypesandtheirthermalsignatures.Developmentofquantitativesurfacedescriptorscouldimproveourcapabilitiesformodelingurbanthermallandscapesandadvanceurbanclimateresearch.ThisstudydevelopedananalyticalprocedurebaseduponaspectralunmixingmodelforcharacterizingandquantifyingtheurbanlandscapeinIndianapolis,Indiana.ALandsatEnhancedThematicMapperPlusimageofthestudyarea,acquiredon22June2002,wasspectrallyunmixedintofourfractionendmembers,namely,greenvegetation,soil,highandlowalbedo.Impervioussurfacewasthencomputedfromthehighandlowalbedoimages.Ahybridclassificationprocedurewasdevelopedtoclassifythefractionimagesintosevenland-useandland-coverclasses.Next,pixel-basedLSTmeasurementswererelatedtourbansurfacebiophysicaldescriptorsderivedfromspectralmixtureanalysis(SMA).Correlationanalyseswereconductedtoinvestigateland-coverbasedrelationshipsbetweenLSTandimpervioussurfaceandgreenvegetationfractionsforananalysisofthecausesofLSTvariations.ResultsindicatethatfractionimagesderivedfromSMAwereeffectiveforquantifyingtheurbanmorphologyandforprovidingreliablemeasurementsofbiophysicalvariablessuchasvegetationabundance,soil,andimpervioussurface.AnexaminationofLSTvariationswithincensusblockgroupsandtheirrelationshipswiththecompositionsofLULCtypes,biophysicaldescriptors,andotherrelevantspatialdatashowsthatLSTpossessedaweakerrelationwiththeLULCcompositionsthanwithothervariables(includingurbanbiophysicaldescriptors,remotesensingbiophysicalvariables,CIS-basedimpervioussurfacevariables,andpopulationdensity).Furtherresearchshouldbedirectedtorefinespectralmixturemodeling.Theuseofmulti-temporalremotesensingdataforurbantime-spacemodelingandcomparisonofurbanmorphologyindifferentgeographicalsettingsarealsofeasible.DOI:10.14358/PERS.72.11.1275CrossrefPDF
|
| [33] | WengQihao,LuDengsheng,SchubringJ.Estimationoflandsurfacetemperature-vegetationabundancerelationshipforurbanheatislandstudies.RemoteSensingofEnvironment,2004,89(4):467-483.Remotesensingofurbanheatislands(UHIs)hastraditionallyusedtheNormalizedDifferenceVegetationIndex(NDVI)astheindicatorofvegetationabundancetoestimatethelandsurfacetemperature(LST)-vegetationrelationship.Thisstudyinvestigatestheapplicabilityofvegetationfractionderivedfromaspectralmixturemodelasanalternativeindicatorofvegetationabundance.ThisisbasedonexaminationofaLandsatEnhancedThematicMapperPlus(ETM+)imageofIndianapolisCity,IN,USA,acquiredonJune22,2002.ThetransformedETM+imagewasunmixedintothreefractionimages(greenvegetation,drysoil,andshade)withaconstrainedleast-squaresolution.Thesefractionimageswerethenusedforlandcoverclassificationbasedonahybridclassificationprocedurethatcombinedmaximumlikelihoodanddecisiontreealgorithms.ResultsdemonstratethatLSTpossessedaslightlystrongernegativecorrelationwiththeunmixedvegetationfractionthanwithNDVIforalllandcovertypesacrossthespatialresolution(30to960m).Correlationsreachedtheirstrongestatthe120-mresolution,whichisbelievedtobetheoperationalscaleofLST,NDVI,andvegetationfractionimages.Fractalanalysisofimagetextureshowsthatthecomplexityoftheseimagesincreasedinitiallywithpixelaggregationandpeakedaround120m,butdecreasedwithfurtheraggregation.ThespatialvariabilityoftextureinLSTwaspositivelycorrelatedwiththoseinNDVIandinvegetationfraction.TheinterplaybetweenthermalandvegetationdynamicsinthecontextofdifferentlandcovertypesleadstothevariationsinspectralradianceandtextureinLST.Thesevariationsarealsopresentintheotherimagery,andareresponsibleforthespatialpatternsofurbanheatislands.Itissuggestedthatthearealmeasureofvegetationabundancebyunmixedvegetationfractionhasamoredirectcorrespondencewiththeradiative,thermal,andmoisturepropertiesoftheEarth'ssurfacethatdetermineLST.DOI:10.1016/j.rse.2003.11.005ElsevierPDF
|
| [34] | WangZhengxing,LiuChuang,HueteAlfredo.FromAVHRR-NDVItoMODIS-EVI:Advancesinvegetationindexresearch.ActaEcologicaSinica,2003,23(5):979-987.GlobalAVHRR-NDVIdatasetshavebeenwidelyappliedtomanyfieldsfromlandcoverchangetotheextractionofvariousbiophysicalvegetationparameters.YettherestillremainsomelimitationsintheNDVIproduct:(1)NDVIsaturatesinwell-vegetatedareas,partlyaresultofsaturationintheRedchannelandpartlyduetotheratio-basedNDVIequation;(2)TheeffectofcanopybackgroundonNDVIhasnotbeenconsidered;(3)TheratioingpropertiesoftheNDVIalongwiththeMaximumValueComposite(MVC)proceduredoesremovesomesourcesofinternalandexternalnoise,buttherestillremainsignificantnoiseinthefinalNDVIproducts;(4)TheMVCcannotguaranteetheselectionoftheclearestpixelsandsmallestviewangles.AlloftheselimitationsareimprovedtosomeextentintheEnhancedVegetationIndex(EVI)productfromtheModerateResolutionImagingSpectroradiometer(MODIS).TheMODIS-EVIhasseveraladvantagesovertheAVHRR-NDVI;(1)TheMODISatmospherecorrectionschemeincludestheeffectofatmosphericgases,aerosol,thincirrusclouds,watervapor,andozone,whereasthereareonlycorrectionsforRayleighscatteringandozoneabsorptionintheAVHRR-NDVIproduct.Thisreducestheneedforratio-basedvegetationindices,suchastheNDVI,thatremovesomeatmosphericnoiseatthecostofsaturation;(2)TheinfluenceofresidualaerosolisremovedbytheAtmosphereResistantVegetationIndex(ARVI),whichisbasedonthedifferenceofRedandBlueaerosolscattering;(3)TheinfluenceofthecanopybackgroundisreducedbytheSoilAdjustedVegetationIndex(SAVI);(4)TheconceptsbehindtheARVIandSAVIarecoupledtogethertoformtheEnhancedVegetationIndex(EVI),whichremovesbothatmosphereandbackgroundnoisesimultaneouslyand;(5)AConstrained-ViewMaximumValueComposite(CV-MVC)algorithmisappliedtoselecttheclearestpixelswithsmallestviewanglesandaBRDFcompositingschemeisbeingtestedtofurtherimprovetheseasonaldepictionofvegetationdynamics.TheMODIS-EVIhasimproveditslinearitywithvegetation,particularlyinwell-vegetatedregions.DOI:10.1023/A:1022289509702CNKI[王正兴,刘闯,HueteAlfredo.植被指数研究进展:从AVHRR-NDVI到MODIS-EVI.生态学报,2003,23(5):979-987.]万方
|
| [35] | HueteAR,JusticeCO,VanLeeuwenW.MODISvegetationindex(MOD13).Version3.Algorithmtheoreticalbasisdocument.GreenbeltMD:NASA,GoddardSpaceFlightCenter,1999:44-45.esearchGat
|
| [36] | ChengLiyu,ZhouYi,WangLitao,etal.AnestimateofthecitypopulationinChinausingDMSPnight-timesatelliteimagery.IGARSS2007,2007:691-694.esearchGat
|
| [37] | SuttonPC,TaylorMJ,ElvidgeCD.UsingDMSPOLSimagerytocharacterizeurbanpopulationsindevelopedanddevelopingcountries.RemoteSensingofUrbanandSuburbanAreas,2010:329-348.esearchGat
|
| [38] | ChenJin,ZhuoLi,ShiPeijun,etal.ThestudyonurbanizationprocessinChinabasedonDMSP/OLSdata:Developmentofalightindexforurbanizationlevelestimation.JournalofRemoteSensing,2003,7(3):168-175,241.Urbanization,stimulatedbystrikingeconomicdevelopment,hasbeenproceededinChinaonalargescaleandwithstrikingrapidityinthepasttwodecades.ItisnecessarytomonitorandmodelurbanizationprocessofChinaforitssustainabledevelopment.ThispaperpresentsanewlightindexforregionalurbanizationlevelestimationconsideringthelightspatialdistributionandintensitybasedonDMSP/OLSdata,whichwaspre-processedbyJapanNationalInstituteofEnvironmentalStudies.Thecorrelationanalysisbetweenlightindexandcompositeurbanizationindexwascarriedoutinprovincescale.Theresultshowsthatthereissignificantrelationshipbetweentwoindexes.TheregressionmodelforcompositeurbanizationindexestimationusinglightindexwasalsodevelopedwithR2equalto0.793.Itsuggeststhatlightindexisaneffectiveandapplicableindexforregionalurbanizationanalysisandmonitoring.ThroughtheanalysisoflightindexchangeinChinaduring1992to1998,itisshownthattheurbanizationlevelinChinaisdifferentfromhighlevelinEastChinatolowlevelinWestChina,andurbanizationlevelwasimprovedlargelyduring1992to1998,especiallyinprovincesofEastChina.CNKI[陈晋,卓莉,史培军,等.基于DMSP/OLS数据的中国城市化过程研究:反映区域城市化水平的灯光指数的构建.遥感学报,2003,7(3):168-175,241.]万方
|
| [39] | ZhuoLi,ShiPeijun,ChenJin,etal.ApplicationofcompoundnightlightindexderivedfromDMSP/OLSdatatourbanizationanalysisinChinainthe1990s.ActaGeographicaSinica,2003,58(6):893-902.Thispaperpresentedcompoundednightlightindex(CNLI)derivedfromDMSP/OLSdatawhichwerepre-processedbyJapanNationalInstituteofEnvironmentalStudiesforregionalurbanizationlevelestimationconsideringthelightspatialdistributionandintensity.Thecorrelationanalysesbetweenlightindicesandcompositeurbanizationindiceswerecarriedoutatprovinciallevelandcountylevel.Theycanalsobecarriedoutinotherscalesifappropriatecensusdataareavailable.Theresultsshowthatthereweresignificantre-lationshipsbetweenthetwokindsofindices.TheregressionmodelsatprovincialscaleandcountyscaleforcompositeurbanizationindexestimationusingCNLIwerealsodeveloped.ItsuggeststhatCNLIisaneffectiveandapplicableindexforregionalurbanizationanalysisandmonitoring.AnalysisofthechangesofCNLI,SandIinChinaduringtheperiod1992-1998showedthattheimbalanceofurbanizationlevelinChinawasobvious.Itwashigherintheeastandlowerinthewest.Urbanizationlevelwasimprovedlargelyduring1992-1998,butthespeedandthetypeofdevelopmentweredifferent.原文[卓莉,史培军,陈晋,等.20世纪90年代中国城市时空变化特征:基于灯光指数CNLI方法的探讨.地理学报,2003,58(6):893-902.]原文
|
| [40] | ZhangQingling,SetoKC.Mappingurbanizationdynamicsatregionalandglobalscalesusingmulti-temporalDMSP/OLSnighttimelightdata.RemoteSensingofEnvironment,2011,115(9):2320-2329.Urbanareasconcentratepeople,economicactivity,andthebuiltenvironment.Assuch,urbanizationissimultaneouslyademographic,economic,andland-usechangephenomenon.Historically,theremotesensingcommunityhasusedopticalremotesensingdatatomapurbanareasandtheexpansionofurbanland-coverforindividualcities,withlittleresearchfocusedonregionalandglobalscalepatternsofurbanchange.However,recentresearchindicatesthaturbanizationatregionalscalesisgrowinginimportanceforeconomics,policy,landuseplanning,andconservation.Therefore,thereisanurgentneedtounderstandandmonitorurbanizationdynamicsatregionalandglobalscales.Here,weillustratetheuseofmulti-temporalnighttimelight(NTL)datafromtheU.SAirForceDefenseMeteorologicalSatellitesProgram/OperationalLinescanSystem(DMSP/OLS)tomonitorurbanchangeatregionalandglobalscales.Weuseindependentlyderiveddataonpopulation,landuseandlandcovertotesttheabilityofmulti-temporalNTLdatatomeasureregionalandglobalurbangrowthovertime.Weapplyaniterativeunsupervisedclassificationmethodonmulti-temporalNTLdatafrom1992to2008tomapurbanizationdynamicsinIndia,China,Japan,andtheUnitedStates.Fortwo-yearintervalsbetween1992and2000,IndiaconsistentlyexperiencedhigherratesofurbangrowththanChina,andbothcountriesexceededtheurbangrowthratesoftheUnitedStatesandJapan.ThisisnotsurprisinggiventhatthepopulationsofIndiaandChinaweregrowingfasterthanthoseoftheU.S.andJapanduringthoseperiods.Fortwo-yearintervalsbetween2000and2008,ChinaexperiencedhigherratesofurbangrowththanIndia.Resultsshowthatthemulti-temporalNTLprovidesaregionalandpotentiallyglobalmeasureofthespatialandtemporalchangesinurbanizationdynamicsforcountriesatcertainlevelsofGDPandpopulation-drivengrowth.DOI:10.1016/j.rse.2011.04.032Elsevier
|
| [41] | LoCP,WelchR.Chineseurbanpopulationestimates.AnnalsoftheAssociationofAmericanGeographers,1977,67(2):246-253.
|
Full-Text
|
|
Contact Us
service@oalib.com QQ:3279437679 
WhatsApp +8615387084133
|
|
