본문 바로가기
KERD
논문KCI 등재

머신러닝 기법과 분광지수를 활용한 산불피해 강도 분류 - 2023년 충청남도 홍성군 산불피해지를 대상으로 -

Classification of Wildfire Severity Using Machine Learning Algorithms and Spectral Indices - A Case Study of 2023 Hongseong Wildfire -

이영진|이상인|임채준
(사)위기관리이론과실천 (Crisis and Emergency Management: Theory and Praxis)|2024.12|Crisisonomy|제20권 제12호|pp. 121-137

0

조회

0

다운로드

0

인용

0

참고문헌

키워드

머신러닝machine learningSentinel-2분광지수산불피해 강도 분류spectral indiceswildfire severity classification

초록

본 연구는 2023년 충청남도 홍성군 산불피해지 대상으로 Maximum Likelihood Classification(MLC), Random Forest(RF), Support Vector Machine(SVM), K-Nearest Neighbors(KNN) 이상 4가지 머신러닝 알고리즘 모델의 산불피해 강도 분류 성능을 파악하기 위해 실시되었다. 검증데이터와 비교한 결과, 모델 총 성능은 Random Forest(RF)의 Overall Accuacy(OA)와 Cohen’s Kappa 계수가 97.2%와 0.95로 가장 높은 성능을 보였다. RF는 여러 개의 의사결정 나무들을 결합하여 최종 예측을 하므로 데이터의 불균형을 해소할 수 있다. 또한, 산불피해 강도에 따른 성능은 모든 머신러닝 알고리즘 모델에서 열해 피해지가 수관화 및 지표화 피해지에 비해 낮은 탐지 성능을 나타냈다. 이는 수관화ㆍ열해 피해지의 분광 반사율 패턴이 유사하고, 열해·지표화 피해지의 경우 복잡한 수관층 구조로 인해 탐지 성능이 낮은 것으로 판단된다. 이와 같은 머신러닝 알고리즘 모델을 통한 산불피해 강도별 면적을 산출은 신속한 복구계획 수립뿐만 아니라 산불로 인한 온실가스 배출량 산정에 기초자료를 제공할 수 있을 것이다.

Abstract

This study aims to evaluate the wildfire severity classification performance of four machine learning algorithm models, Maximum Likelihood Classification(MLC), Random Forest(RF), Support Vector Machine(SVM) and K-Nearest Neighbors(KNN), in the wildfire areas of Hongseong-gun, Chungcheongnam-do, in 2023. Compared with the verification dataset, the overall performance of the models indicated that the Random Forest(RF) showed the highest accuracy. The highest performance of the Random Forest(RF) model can be explained by its ability to combine multiple decision trees for a final prediction. Furthermore, the performance based on wildfire severity showed that all machine learning algorithm models had lower detection performance in heat damage areas compared with crown and surface fire areas. Crown and heat damage areas have similar spectral reflectance patterns, and heat damage and surface fire areas are considered to have shown lower performance due to the complex canopy structure. Machine learning based burn severity estimation provide critical baselines for rapid recovery planning and greenhouse gas emissions calculations.

참고문헌 (53)

  1. Bishop, C. M., "Pattern Recognition and Machine Learning" (2006), Springer
  2. Chen, J., "The Effects of Spatial Resolution and Resampling on the Classification Accuracy of Wetland Vegetation Species and Ground Objects: A Study Based on High Spatial Resolution UAV Images", Drones, 7(1), pp.61~ (2023)
  3. Collins, L., "Training Data Requirements for Fire Severity Mapping Using Landsat Imagery and Random Forest", Remote Sensing of Environment, 245, pp.111839~ (2020)
  4. "2023:A Year of Intense Global Wildfire Activity" (2023)[URL]
  5. European Space Agency, "Level-1 Algorithm Theoretical Bases Document", 8~12 (2023)
  6. "Maximum Likelihood Classification" (2024)[URL]
  7. Filipponi, F., "BAIS2 : Burned Area Index for Sentinel-2", The 2nd International Electronic Conference on Remote Sensing, pp.364~ (2018)
  8. Ghayour, L., "Performance Evaluation of Sentinel-2 and Landsat 8 OLI Data for Land Cover", Use Classification Using a Comparison between Machine Learning Algorithms, 2021, pp.13(7), 1349~
  9. Ghimire, B., "An Evaluation of Bagging, Boosting, and Random Forests for Land-cover Classification in Cape Cod, Massachusetts, USA", GIScience & Remote Sensing, 49(5), pp.623~643 (2012)
  10. Gholamnia, K., "Comparisons of Diverse Machine Learning Approaches for Wildfire Susceptibility Mapping", Symmetry, 12(4), pp.604~ (2020)
  11. Gibson, R., "A Remote Sensing Approach to Mapping Fire Severity in South-eastern Australia Using Sentinel 2 and Random Forest", Remote Sensing of Environment, 240, pp.111702~ (2020)
  12. Goodfellow, I., "Deep Learning", pp.69~161 (2016), MIT Press
  13. Intergovermental Panel on Climate Change, "2006 IPCC Guildelines for National Greenhouse Gas Inventories", ~ (2006)
  14. A Review of Assessing the Accuracy of Classifications of Remotely Sensed Data and of Methods Including Remote Sensing Data in Forest Inventory, "International Institute for Applied Systems Analysis", ~ (1998)
  15. Jensen, J. R., "Biophysical Remote Sensing", Annals of the Association of American Geographers, 73(1), pp.111~132 (1983)
  16. Kai, P. M., "Effects of Resampling Image Methods in Sugarcane Classification and the Potential Use of Vegetation Indices Related to Chlorophyll", 2021 IEEE 45th Annual Computers, Software, and Applications Conference (COMPSAC), pp.1526~1531 (2021)
  17. Kang, Kyoung Hee, "Study on the Effect of Training Data Sampling Strategy on the Accuracy of the Landslide Susceptibility Analysis Using Random Forest method", Economic and Environmental Geology, 52(2), pp.199~212 (2019)
  18. Key, C. H., "FIREMON: Fire Effects Monitoring and Inventory System", pp.1~55 (2006)
  19. Kim, Bo Mi, "A Study on the Optimal Classification Method for Forest Fire Burn Severity Based on Remote Sensing", Ph. D., pp.2021
  20. Kim, Byeong Cheol, "Forest Burned Area Detection Using Landsat 8", 9 and Sentinel-2 A, Korean Journal of Remote Sensing, pp.2022, 38[DOI]
  21. Kim, Dae Sun, "Classification of Wildfire Burn Severity Using the Clustering Methods with Sentinel-2 Images", Journal of Climate Research, 15(3), pp.173~185 (2020)
  22. Kim, Eung Nam, "An Investigation of Pine wilt Damage by Using Ground Remote Sensing Technique", Journal of the Korean Association of Regional Geographers, 14(1), pp.84~92 (2008)
  23. A Study on Damage Characteristics and Development of Burn Severity Evaluation Methods, "Korea Forest Research Institute", 37~96 (2013)
  24. Annual Report on Forest Fire Statistics, "Korea Forest Service", 53~192 (2024)
  25. Landis, J. R., "The Measurement of Observer Agreement for Categorical Data", Biometrics, 33(1), pp.159~174 (1977)
  26. Lee, Han Na, "Mapping Burned Forests Using a k-Nearest Neighbors Classifier in Complex Land Cover", KSCE Journal of Civil and Environmental Engineering Research, 43(6), pp.883~896 (2023)
  27. Lee, Jong Ook, "Analysis of the Potential Landslide Hazard after Wildfire Considering Compound Disaster Effect", Journal of the Korea Society of Environmental Restoration Technology, 22(1), pp.33~45 (2019)
  28. Lee, Soo Jin, "Development of FBI(Fire Burn Index)for Sentinel-2 Images and an Experiment for Detection of Burned Areas in Korea", Journal of the Association of Korean Photo-Geographers, 27(4), pp.187~202 (2017)
  29. Lee, Ye Eun, "Estimation of Carbon Dioxide Emission from a Fires with Burn Severity", Crisisonomy, 19(6), pp.89~100 (2023)
  30. Lin, W., "An Ensemble Random Forest Algorithm for Insurance Big Data Analysis", Ieee Access, 5, pp.16568~16575 (2017)
  31. Liu, M., "Comparison of Random Forest, Support Vector Machine and Back Propagation Neural Network for Electronic Tongue Data Classification : Application to the Recognition of Orange Beverage and Chinese Vinegar", Sensors and Actuators B: Chemical, 177, pp.970~980 (2013)
  32. Montesinos López, O. A., "Multivariate Statistical Machine Learning Methods for Genomic Prediction", Springer Nature, pp.109~140 (2022)
  33. Montorio, R., "Unitemporal Approach to Fire Severity Mapping Using Multispectral Synthetic Databases and Random Forests", Remote Sensing of Environment, 249, pp.112025~ (2020)
  34. Müller, A. C., "Introduction to Machine Learning with Python: A Guide for Data Scientists" (2016), O'Reilly Media
  35. Reporting on Wildfire Damage Analysis Using Satellite Imagery, "National Institute of Forest Science", ~ (2023)
  36. Ponomarev, E. I., "Wildfire Intensity and Fire Emissions in Siberia", Fire, 6(7), pp.246~ (2023)
  38. Rouse, J. W., "Monitoring Vegetation Systems in the Great Plains with ERTS", Earth Resources Technology Satellite, 3(20), pp.309~318 (1974)
  39. Roy, D. P., "Remote Sensing of Fire Severity: Assessing the Performance of the Normalized Burn Ratio", IEEE Geoscience and Remote Sensing Letter, 3(1), pp.112~116 (2006)
  40. Sarica, A., "Random Forest Algorithm for the Classification of Neuroimaging Data in Alzheimer's Disease: A Systematic Review", Frontiers in Aging Neuroscience, 9, pp.329~ (2017)
  41. Seydi, S. T., "Wildfire Damage Assessment over Australia Using Sentinel-2 Imagery and MODIS Land Cover Product within the Google Earth Engine Cloud Platform", Remote Sensing, 13(2), pp.220~ (2021)
  42. Shin, Jugn Ii, "Analysis of Availability of High-resolution Satellite and UAV Multispectral Images for Forest Burn Severity Classification", Korean Journal of Remote Sensing, 35(6), pp.1095~1106 (2019)
  43. Sim, Sung Mum, "Wildfire Severity Mapping Using Sentinel Satellite Data Based on Machine Learning Approaches", Korean Journal of Remote Sensing, 36(5), pp.1109~1123 (2020)
  44. Stephen, J. P., "Introduction to Wildfire", pp.92~127 (1996), Wiley
  45. Volkova, L., "Tree Mortality and Carbon Emission As a Function of Wildfire Severity in South-eastern Australian Temperate Forests", Science of the Total Environment, 853, pp.158705~ (2022)
  46. Won, Myong Soo, "Analysis of Burn Severity in Large-fire Area Using SPOT5Images and Field Survey Data", Korean Journal of Agricultural and Forest Meteorology, 16(2), pp.114~124 (2014)
  47. Won, Myong Soo, "An Quantitative Analysis of Severity Classification and Burn Severity for the Large Forest Fire Areas Using Normalized Burn Ratio of Landsat Imagery", Journal of the Korean Association of Geographic Information Studies, 10(3), pp.80~92 (2007)
  48. Won, Myong Soo, "Estimation on Greenhouse Gases (GHGs) Emission of Large Forest Fire Area in 2013", Journal of the Korean Association of Geographic Information Studies, 17(3), pp.54~67 (2014)
  49. Wu, M., "A Remote Sensing Spectral Index Guided Bitemporal Residual Attention Network for Wildfire Burn Severity Mapping", IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17, pp.17187~17206 (2024)
  50. Yang, Min Sun, "A Study on Estimation of Forest Burn Severity Using Kompsat-3A Images", Korean Journal of Remote Sensing, 39(6), pp.1299~1308 (2023)
  51. Youn, Hyoung Jin, "Classification of Forest Fire Damage Grade Using Machine Learning and Sentinel-2", The Korea Spatial Planning Review, 106, pp.107~117 (2020)
  52. Zhu, Z., "Improvement and Expansion of the Fmask Algorithm : Cloud, Cloud Shadow, and Snow Detection for Landsats 4-7, 8, and Sentinel 2 Images", Remote Sensing of Environment, 159, pp.269~277 (2015)
  53. Zikiou, N., "Remote Sensing and Machine Learning for Accurate Fire Severity Mapping in Northern Algeria", Remote Sensing, 16(9), pp.15~17 (2024)

관련 논문AI추천

기계학습을 이용한 Sentinel-2 산불피해등급 분류윤형진, 정종철국토연구원 · 국토연구2020 · 제106권 랜덤포레스트와 임상도를 사용한안동 산불 피해 범위 분석손주형, 정종철한국국토정보공사 · 지적과 국토정보2025 · 제55권 제1호Sentinel-2 영상과 클러스터링 기법을 이용한 산불피해강도 분류 - 2020년 안동 산불을 사례로 -김대선, 이양원기후연구소 · 기후연구2020 · 제15권 제3호Sentinel-2A 영상과 현장자료를 이용한 홍성 산불 피해강도 구분 및온실가스 배출량 산정박라혜, 서연옥, 윤나경, 이영진, 임채준농업생명과학연구원 · 농업생명과학연구2025 · 제59권 제5호고해상도 원격탐사 자료와 기계학습을 이용한 한국 산림의탄소 저장량 산정신재원, 장동영, 정수종한국기상학회 · 대기2023 · 제33권 제1호머신러닝과 샘플링을 이용한 강원도 지역 산불발생예측모형 개발박지현, 이유리, 조용주, 채경재사)한국빅데이터학회 · 한국빅데이터학회 학회지2018 · 제3권 제2호

이 간행물의 다른 논문 (14건)

극단치 분석 및 극한 현상 모델링에 관한 연구남호수, 신희건pp. 1-13에너지 정책 네트워크에서 정책행위자 간 관계구조와 상호작용 분석 독일과 한국의 시민참여 사례분석을 통한 함의 찾기박지민, 김서용, 노유경, 전세혁, 이시현, 임진영pp. 15-35노후 댐의 안전 및 유지관리 개선을 위한 신뢰성 평가 기법문장원, 최병한, 권현한pp. 37-47향정신성의약품 오·남용에 관한 연구 동향김영호, 김혜선, 이연재, 라세원pp. 49-65지방소멸 대응을 위한 사회안전에 영향을 미치는 위험요인 규명에 관한 연구 - 충청남도 소멸고위험지역과 보통지역의 20~39세 여성을 대상으로 -박근오pp. 67-80
원문 보기