评估易受灾害影响的斐济的新损害评估技术（英）
AI智能总结
NO.240 FEBRUARY2023 KEYPOINTS •Recentadvancesintechnologymakesatelliteimageryanimportantsourceofinformationforeffectivedisasterriskmanagement ADBBRIEFS AppraisingNewDamageAssessmentTechniquesinDisaster-ProneFiji andpost-disasterdamageassessment.Theimpactoftropicalcycloneson agricultureinFijiisestimatedbymeasuringtheEnhancedVegetationIndexderivedfromsatelliteimages, beforeandaftertheevent,andlinkedwithavailablehouseholdsurveysandagriculturalcensusdatatoassesseconomicimpact. •Theanalysispavesapossibleapproachforestimating thepredictedimpactof IlanNoy Professor SchoolofEconomicsandFinanceVictoriaUniversityofWellingtonWellington,NewZealand HomerPagkalinawanAssociateEconomicsOfficerEconomicResearchandRegional CooperationDepartmentAsianDevelopmentBank MadhaviPundit SeniorEconomist EconomicResearchandRegionalCooperationDepartment AsianDevelopmentBank PriscilleVillanueva EconomicsOfficer EconomicResearchandRegionalCooperationDepartment AsianDevelopmentBank atropicalcycloneontheagriculturesectorofacyclone-proneregion,basedonlyonthecyclone’sknowntrajectory,andevenbeforetheavailabilityofpost-eventsatelliteimagery.Themethodcanbecomeacomponent ina“nowcastingtool”toassesstheimmediateimpactofadisaster.Theimpact andneedsassessmentcanberefinedasimagesandsurveyinformationbecomeaccessibleovertime. •Moreworkisneededtoimprovethepreliminaryestimatesandapproachprimarilybyaccessinggranularanddetaileddatasets,includingonagriculturecropsandhouseholdincomes. ISBN978-92-9270-023-2(print) ISBN978-92-9270-024-9(electronic) ISSN2071-7202(print) ISSN2218-2675(electronic)PublicationStockNo.BRF230028-2 DOI:http://dx.doi.org/10.22617/BRF230028-2 ADVANCESINPOST-DISASTERDAMAGEASSESSMENT Recenttechnologicaladvancesinremotesensinghaveledtosignificantincreasesintemporal,spatial,andspectralresolutionsoftheseimages.Improvementshavealsobeenmadeindataavailabilityandaccessibility,aswellasadvancesintheprocessingmethodsrequiredtoeffectivelyinterprettheselargeamountsofdata. Thesedevelopmentshaveledtoanincreaseintheopportunitiesfortheuseofsatelliteimageryformoreeffectivedisasterriskmanagement. Onesuchapplicationthathasadvancedbecauseoftheseimprovementsispost-disasterdamageassessment.Traditionally,inapost-disasterscenario,ateamofexpertsissent totheimpactareatoconductagroundsurveyandassessdamages(oftenconnectedtoaPost-DisasterNeedsAssessmentreport),whichtheAsianDevelopmentBank(ADB)typicallyundertakesjointlywithimpactedcountrygovernmentsandotherrelevantagencies.Whileextremelyuseful,thisprocessistime-consuming,costly,andproblematic becauseoftenaccesstotheaffectedareaisdifficult,dangerous,orrestricted.Alternatively,governments,andespeciallypublicandprivateinsurancecompanies,useriskmodelstoquantifythedamages.Riskmodels,however,canbeinaccurateandaregenerallyunable toaccountforcompoundingorcascadingevents,whicharemuchmoredifficulttoquantitativelymodel.Similarly,thestandardapproachtonowcastdisasterimpacts,whichreliesonriskmodels,doesnottypicallyaccountforthecompoundingimpactofvarioushazardphenomena,suchaswindandrainfallassociatedwithtropicalcyclones. ThegrantfundforthestudywasreceivedfromtheJapanFundforProsperousandResilientAsiaandthePacificfinancedbytheGovernmentofJapanthroughtheAsianDevelopmentBank. Satelliteimagerymayprovideaneasilyavailableandaccuratedatasourcetogaugeadisaster’sspecificimpactsandcanaccountfor adisaster’scompoundingeffects.Whileitcanserveasastrongcomplement,ifaccurateenough,satelliteimagerymaypotentiallyreplacesomecomponentsofthegroundsurveys.Anapproachthathasbeencalibratedwithremote-sensingimagerycanalsobeusedasacomponentinanowcastingtooltoassesstheimpactofaspecificdisaster,forexample,atropicalcyclone,basedonlyonitsknowntrajectory,andevenbeforepost-eventsatelliteimageryisavailable.Wediscusssuchanapproachinthisbrief. ASSESSINGIMPACTOFTROPICALCYCLONESONFIJI’SAGRICULTURE Aknowledgeandtechnicalassistanceiscurrentlyunderwaytoinvestigatethefeasibilityofusinginnovativedata,suchassatelliteimageryfornowcastingandpost-disasterdamageassessment.1Noyetal.(2022)examinethepossibilityofcombiningremote- sensingdatawithsocioeconomicsourcesofinformation(householdsurveysandthecensus)tonowcastthedamagefromtropicalcyclonesonagriculture.Indoingso,itgoesbeyondthenow-routineprocedureoflinkinghazardindicators(suchasstrongwinds)withremote-sensingdata,tothenextstepofassessingeconomicimpacts. Inthisstudy,dataonvegetationcoverderivedfromsatelliteimagesbeforeandaftertropicalcyclonesinFijiarelinkedwithavailablehouseholdsurveysandtheagriculturalcensusdatatoobtain animprovedassessmentoftropicalcycloneimpacts.Tropicalcyclonesarethemaindisaster-inducinghazardinFijiandinmanyPacificislandcountries(PICs).Damagesareassessedaroundfourdestructivetropicalcyclones(TCs)thatoccurredinFijibetween2016and2020.Theyare •TCWinston(15February2016), •TCsJosie(2April2018)andKeni(10April2018),and •TCHarold(7–8April2020). Asmeasuredbywindspeed,TCWinston(aCategory5cyclone)wasoneofthestrongesttropicalstormsinrecordedhistory andthestrongestcycloneevertomakelandfallintheSouthernHemisphere,withestim
