外文翻译以GIS为基础的煤矿开采沉陷危机分析.docx

外文翻译以GIS为基础的煤矿开采沉陷危机分析.docx

《外文翻译以GIS为基础的煤矿开采沉陷危机分析.docx》由会员分享，可在线阅读，更多相关《外文翻译以GIS为基础的煤矿开采沉陷危机分析.docx（16页珍藏版）》请在冰豆网上搜索。

外文翻译以GIS为基础的煤矿开采沉陷危机分析

翻译部分

英文原文

SensitivityanalysisfortheGIS-basedmappingofthegroundsubsidencehazardnearabandonedundergroundcoalmines

AbstractGroundsubsidencearoundabandonedunder-groundcoalminescancausemuchlossoflifeandprop-erty.WeanalyzefactorsthatcanaffectgroundsubsidencearoundabandonedminesinJeongahminKangwon-dobysensitivityanalysisingeographicinformationsystem（GIS）.Spatialdataforthesubsidencearea,topographyandgeologyandvariousgroundengineeringdatawerecol-lectedandusedtomakeafactorrasterdatabaseforagroundsubsidencehazardmap.Todeterminetheimpor-tanceofextractedsubsidence-relatedfactors,frequencyratiomodelandsensitivityanalysiswereemployed.Sen-sitivityanalysisisamethodforcomparingthecombinedeffectsofallfactorsexceptone.Sensitivityanalysisanditsverificationshowedthatusingallfactorsprovided91.61%accuracy.Thebestaccuracywasachievedbynotconsid-eringthegroundwaterdepth（92.77%）andtheworstbynotconsideringthelineament（85.42%）.Theresultsshowthatthedistancefromthelineamentandthedistancefromthedrifthighlyaffectedtheoccurrenceofgroundsubsidence,andthegroundwaterdepth,landuseandrockmassrating

hadtheleasteffects.Thus,wedeterminedcausesofgroundsubsidenceinthestudyareaandthisinformationcouldhelpinthepredictionofgroundsubsidenceinotherareas.

KeywordsGISAbandonedundergroundcoalmineGroundsubsidenceFrequencyratioSensitivityanalysisKorea

Introduction

CoalindustryhasplayedamajorroleinthedevelopmentoftheKoreaneconomysincethe1960s.However,thecoalindustrybegantodeclinefromtheendofthe1980sbecausetheoilpricefellglobally.Therewere173coalminesin1988inJeongseon,Kangwon-do,thelargestcoalminingregioninKorea.However,valuelesscoalmineswereabandonedaftertheimplementationofcoalindustryrationalizationactionbyKoreangovernment.Today,nocoalminesareinoperation.IntheabandonedcoalminesinJeongseon,environmentalpollution,havingneverbeenconsideredafutureproblematthetimetheminesceasedoperating,isontheincrease.Heavymetalsinwastewaterfromcoalminesandmuckreachingriversandsoilisaseriousproblem.Inparticular,mostabandonedcoalminesareinmountainrangesandtheirdamagetotheenviron-mentisnotwidelyknown.Groundsubsidenceduetoanundergroundcoalmineisahazardtohumanlifeaswellastopropertysuchasrailways,roads,housesandbuildings.Inaddition,itisdifficulttocompletelyrecoveranareaofsubsidenceandrehabilitationisexpensive.Moreover,mostcountermeasuresforgroundsubsidencearesimplyrein-forcementsaftergroundsubsidencehasalreadyoccurred.Therefore,itisnecessarytohaveasystematicpredictionandmanagementplanforanareaofgroundsubsidence.Therehavebeenmanystudiesonthepredictionofgroundsubsidencelocationsandhowtodecreaserelateddamage.

Methodology

Ingeneral,topredictgroundsubsidenceitisnecessarytoassumethatsubsidenceoccurrenceisdeterminedbysub-sidence-relatedfactors,andthatfuturesubsidencewilloccurunderconditionsthatarethesameasthoseforpastsubsidenceaccordingtotherelativefrequencyconceptofprobabilitytheory.Onthisbasis,therelationshipbetweenareaswheresubsidencehasoccurredandsubsidence-rela-tedfactorscanbedistinguishedfromtherelationshipbetweenareaswithoutpastsubsidenceandsubsidence-relatedfactors.Inthiswork,thespatialrelationshipsbetweensubsidencelocationsandeachsubsidence-relatedfactorwereanalyzedusingthefrequencyratiowhichisoneoftheprobabilitymodels.Thefrequencyratioistheratiooftheareawheresubsidencehasoccurredtothetotalstudyarea,andisalsotheratiooftheprobabilityofsubsidenceoccurrencetonooccurrenceforagivenattribute.Inthecaseofsubsidenceoccurrence,ifthesubsidenceeventisdenotedBandagivenfactor’sattributeisdenotedD,thenthefrequencyratioofDistheratiooftheconditionalprobabilitiesofB.Therefore,thegreaterthisratio,thestrongertherelationshipbetweensubsidenceoccurrenceandthegivenfactor’sattribute.Thefrequencyratiowascalculatedforeachfactortypeorrangethathadbeenidentifiedassignificantwithrespecttocausinggroundsubsidence.Theratioswereusedforcalculatingthegroundsubsidencehazardindex（GSHI）andmapping.TherehavebeenafewstudiescarriedoutongroundsubsidencehazardevaluationsusingthefrequencyratioandaGIS;forexample,Kimetal.（2006）introducedtheanalysisofgroundsubsidencehazardandsummarizedmanygroundsubsidencehazardevolutionstudies.

DiagramspresentedinFig.4showthestepsforcalcu-latingthefrequencyratio:

（a）findinglocationsofgroundsubsidence,（b）representingcellsofclass1forfactorAand（c）describingtheareaofspatialoverlapforareasofsubsidenceandareasofclass1forfactorA.Inthepresent

exampleinwhichthestudyareacomprises64cells,thereare23groundsubsidencecellsand19cellsthatareofclass1forfactorA.Thereare12cellsofsubsidencethatarealsocellsofclass1forfactorA.Thepercentagesforsubsidenceareawithrespecttoclass1forfactorAandtheentiredomainare52.17and29.69%,respectively.Therefore,thefrequencyratioofclass1forfactorAis52.17/29.69%=1.76.

Sensitivityanalysis

Sensitivityanalysesshowhowasolutionchangeswhentheinputfactorsarechanged.Iftheselectionofafactorresultsinarelativelylargechangeintheoutcome,thentheout-comeissaidtodependonthatfactor.Sensitivityanalysisquantifiestheuncertaintyofeachfactor.Thefactorsthathavethegreatestimpactonthecalculatedsubsidencehazardmapcanthereforebeidentified.

Sensitivityanalysesmodelbehaviorbydeterminingtherateofchangeinthemodeloutputasparametersorinputdataarevaried,thusgivinganunderstandingofhowchangesintheinputsinfluencetheoutput.Suchanalysesexplainhowinputsaffecttheoutputandquantifythestrengthsoftheinputsfromthevariationsintheoutput.Sensitivityanalysishasbeenappliedinavarietyofstudies.ArhonditsisandBrett（2005）introducedprocessformula-tionsandsensitivityanalysisforacomplexeutrophicationmodelofLakeWashington,USA.Jooetal.（1998）repre-sentedthesensitivityoftheleachateleveltoreducethehighleachatelevelinlandfill.Sensitivityanalysishasbeenappliedtolandslidesusceptibilityhazardmapping（LeeandTalib2005）.Inthepresentstudy,thesensitivityanalyseswereconductedtodeterminefactorsinvolvedingroundsubsidence.Toextractsubsidence-relatedfactors,thefre-quencyratiomodelandsensitivityanalysiswereemployed.Thesensitivityanalysiscomparedtheeffectsofallfactorsexceptone.

Data

Manystudieshaveidentifiedimportantfactorsthatcon-tributetogroundsubsidencearoundcoalmines（e.g.Waltham1989;CoalIndustryPromotionBoard1997）,includingthedepthandheightofminecavities,excavationmethod,degreeofinclinationoftheexcavation,scopeofmining,structuralgeology,flowofgroundwaterandthemechanicalcharacteristicsrepresentedbytherockmassrating（RMR）.

Inthisstudy,locationsofgroundsubsidenceandfactorsgoverningtheoccurrenceofgroundsubsidencewerecol-lectedinavector-typespatialdatabaseandthenrepre-sentedonagridusingtheArcGISsoftwarepackage.ThespatialdatabaseisshowninTable1.Thedatabaseincludeda1:

5,000groundsubsidencemap,1:

50,000geologicalmap,1:

5,000topographicmap,1:

5,000landusemap,1:

1,200minetunnelmap,boreholedataandsatelliteimagerywith1mresolution.ThesiteJeongahmwaschosenforinvestigationinthisstudy.ReliableaccuracyofthespatialdatabaseisindispensableinaGISenvironment.Forthisreason,accuratemapsauthorizedbynationalorganizationssuchastheCoalIndustryPromotionBoardforgroundsubsidence,theNationalGeographicInforma-tionInstitutefortopographyandlanduse,theMineRec-lamationCorp.forminetunnelsandboreholesandtheKoreaInstituteofGeoscienceandMineralResourcesforgeologywereassembledeventhoughthescalesofthemapsdiffered.

EightvariablesasshowninFig.5,extractedfromtheconstructedspatialdatabase,wereconsideredasfactorsofgroundsubsidencewhencalculatingtheprobability.Con-tours（5mintervals）andsurveybasepointsofelevationwereextractedfromthetopographicmap,andtriangu-latedirregularnetworkwasmadeusingtheelevations.Theangleofslopewasobtainedfromthedigitalelevationmap.Areasofbuildings,mountains,railways,fields,rivers,hybrids,roadsandmiscellaneoususewereextractedfromthelandusemap.Mostliteraturestatesthatthemajorfactoringroundsubsidenceisthescopeofminecavities.Therefore,constructingadatabaseforthedepthsandwidthsofminecavitieswasimportant.Toachievethis,

（1）GPSmeasurementswereusedtodeterminetheexactpositionsofmineheads,

（2）thesepositionswereusedtovectorizeahardcopyoftheminetunnelmapand（3）thevectorizedminetunnelmapwasconvertedtoanASCIIgridfileandsubtractedfromthedigitalelevationmaprasterdata.Therewere13boreholesatthestudysite,butfourboreholesdidnothaveRMRvaluesandgroundwaterlevels.Aninverse-distanceweightinginterpolationwasusedtocontourtheRMRandgroundwaterlevel.Geolog-icaldatawereextractedfromthe1:

50,000geologicalmap.AlineamentwasdetectedfromIKONOSimagerywith1-mresolutionbyastructuralgeologist.Thedistancefromthelineamentwascalculatedat1-mintervalsusingtheEuclideandistancemethod.Thefactorswerethenrepre-sentedby2m92mc