Publications

Detailed Information

Assessment of disaster risks induced by climate change, using machine learning techniques : 머신러닝 기법을 활용한 기후변화 영향에 따른 재해 리스크 평가

DC Field Value Language
dc.contributor.advisor이동근-
dc.contributor.author박상진-
dc.date.accessioned2022-12-29T15:25:13Z-
dc.date.available2022-12-29T15:25:13Z-
dc.date.issued2022-
dc.identifier.other000000173558-
dc.identifier.urihttps://hdl.handle.net/10371/188759-
dc.identifier.urihttps://dcollection.snu.ac.kr/common/orgView/000000173558ko_KR
dc.description학위논문(박사) -- 서울대학교대학원 : 환경대학원 협동과정 조경학, 2022. 8. 이동근.-
dc.description.abstract기후 변화는 우리 세대에게 시급한 위협이다. 자연 재해는 기후 변화로 인해 더 잦은 빈도와 강력하게 발생하고 있어 예측불가성이 커져가고 있다. 특히, 한국의 자연재해는 대부분 기상 현상으로 인해 발생하는데, 지난 10년간 재해로 인한 전체 피해는 주로 태풍(49%)과 호우(40%)에 기인하였다. 따라서 장기적으로 대비하기 위해서는 홍수, 산사태 등 호우와 관련된 위험을 분석하고 평가하는 위험관리가 필요하다.
따라서 본 논문의 주요 연구질문은 다음과 같다: 1) 기후변화로 인한 복잡한 상황에서 다양한 요인을 고려하여 미래의 잠재적 위험을 어떻게 예측할 것인가, 2) 이러한 위험을 줄이기 위해 어떤 노력을 하는 것이 지속가능한가?. 먼저 연안 홍수, 산사태 등 복합적 영향의 미래 위험도를 평가하기 위해 첫째, 최근 연구에서 널리 활용되고 있는 다중 머신러닝(ML) 알고리즘을 확률론적 접근 방식으로 활용하여 현재의 위험도를 분석하였다. 다양한 RCP 기후변화 시나리오 및 지역 기후 모델에 따른 예측 강우량을 고려하여 미래 위험을 추정했습니다. 둘째, 기후변화 영향으로 인한 재난위험 대응을 위한 적응전략의 실효성을 평가하기 위하여, 적응전략으로 중요한 역할을 하는 녹지, 방파제 등 구조적 대책의 효과성과 지속가능성을 여러 적응경로로 나눠 연안침수에 대한 위험저감을 평가하였다.
연구의 결과는 미래의 위험 지역을 식별하고 위험 관리를 위한 의사 결정 과정, 그리고 토지 이용 계획 및 의사 결정 프로세스를 포함한 재난 감소 및 관리 조치에 대해 지원 가능할 것이다.
-
dc.description.abstractClimate change is an urgent threat to our generation. Natural hazards have become more unpredictable, occurring more frequently and with greater force, due to climate change. Natural disasters in Korea are mostly caused by meteorological events. The total damage caused by disasters in the last ten years is attributed mainly to typhoons (49%) and heavy rain (40%). Therefore, risk management, which analyzes and evaluates hazard risk related to heavy rainfall such as flooding and landslides, is needed to prepare for the long term. Also, effective monitoring and detection responses to climate change are critical for predicting and managing threats to hazard risks.
Therefore, the main research questions of this thesis are as follows: 1) How to predict future potential risks in a complex situation due to climate change considering various factors, 2) And what kind of efforts are made to reduce such risks? Is it sustainable? First of all, to assess the future risk of multiple hazards such as coastal flooding, landslide, 1) this study analyzed the present risk by using multiple machine learning (ML) algorithms that have been widely used in recent studies as part of probabilistic approaches, and future risks were estimated by considering the forecasted rainfall according to different representative concentration pathway (RCP) climate change scenarios and regional climate models. Secondly, to evaluate the effectiveness of adaptation strategies to respond to disaster risks posed by climate change impacts, 2) this research analyzed the effectiveness and sustainability of structural measures such as green space and seawall, which are widely used and play an important role as countermeasures against coastal flooding, by dividing into several adaptation pathways.
The results of this study identify future at-risk areas and can support decision-making for risk management and can guide disaster reduction and management measures, including land use planning and decision-making processes.
-
dc.description.tableofcontentsAbstract i
Chapter 1. Introduction 2
1. Background 2
2. Purpose 4
Chapter 2. Prediction of coastal flooding risk under climate change impacts in South Korea using machine learning algorithms 7
1. Introduction 7
2. Materials and Method 9
2.1 Study Area 9
2.2 Machine learning algorithms 10
2.3 Method 11
3. Results 15
3.1 Comparison of ML algorithms 15
3.2 Risk probability map 16
3.3 Future risk under climate change impacts 17
4. Discussion 18
4.1 Regional differences 18
4.2 Significance factor 20
4.3 Methodological implications 21
5. Conclusions 22
Chapter 3. Predicting susceptibility to landslides under climate change impacts in metropolitan areas of South Korea using machine learning 25
1. Introduction 25
2. Materials and Method 28
2.1 Study Area 28
2.2 Data 29
2.3 Landslide factors analysis 30
2.4 Machine learning algorithms and validation 32
2.5 LSA using different algorithms 33
2.6 Predicting landslide susceptibility 34
3. Results 35
3.1 Multi-collinearity and influencing factor analysis 35
3.2 Comparison of machine learning algorithms 37
3.3 Predicting landslide susceptibility 38
4. Discussion 39
4.1 Analysis of results from different ML algorithms 39
4.2 Difference in susceptibilities based on land cover type 40
5. Conclusions 41
Chapter 4. Adaptation strategies to future coastal flooding: performance evaluation of green and grey infrastructure in South Korea 43
1. Introduction 43
2. Materials and Method 46
2.1 Study area 46
2.2 Data 47
2.3 Comparison of machine learning (ML) techniques and coastal flooding risk analysis 49
2.4 Evaluation of coastal flooding risk with ASs 50
2.5 Potential coastal flooding risk depending on different adaptive pathways 51
3. Results 53
3.1 Performances of ML algorithms 53
3.2 Coastal flooding risk with ASs 54
3.3 Potential coastal flooding risk according to different adaptive pathways 56
4. Discussion 59
4.1 Effect of AS according to spatial characteristics 59
4.2 Importance of nature-based solutions as ASs 62
5. Conclusion 63
Chapter 5. Conclusion 66
Bibliography 71
Abstract in Korean 86
-
dc.format.extent87-
dc.language.isoeng-
dc.publisher서울대학교 대학원-
dc.subjectClimatechangeimpacts-
dc.subjectDisasterRiskReduction(DRR)-
dc.subjectRCPscenario-
dc.subjectsealevelchange-
dc.subjectIntegratedCoastalZoneManagement(ICZM)-
dc.subjectLandslidesusceptibility-
dc.subjectNaturebasedSolution(NBS)-
dc.subjectadaptationstrategy-
dc.subject.ddc712.3-
dc.titleAssessment of disaster risks induced by climate change, using machine learning techniques-
dc.title.alternative머신러닝 기법을 활용한 기후변화 영향에 따른 재해 리스크 평가-
dc.typeThesis-
dc.typeDissertation-
dc.contributor.AlternativeAuthorSang Jin Park-
dc.contributor.department환경대학원 협동과정 조경학-
dc.description.degree박사-
dc.date.awarded2022-08-
dc.identifier.uciI804:11032-000000173558-
dc.identifier.holdings000000000048▲000000000055▲000000173558▲-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share