All Issue

2026 Vol.32, Issue 1 Preview Page
Deep Learning-Based Analysis of Wildfire Susceptibility Considering Spatial Autocorrelation
- Focusing on Andong-si and Uiseong-gun, Gyeongsangbuk-do -

공간 교차검증을 적용한 딥러닝 기반 산불 민감성 분석
- 경상북도 안동시와 의성군을 중심으로 -

이현서1 · 안유순2

Lee, Hyun Seo1 · An, Yoo Soon2

1제1저자, 동국대학교 지리학과 석사과정
2교신저자, 동국대학교 지리교육과 조교수
1First Author, Master Student, Department of Geography, Dongguk University
2Corresponding Author, Assistant Professor, Department of Geography Education, Dongguk University
28 February 2026. pp. 81-105
PDF Citation
Abstract
The large-scale wildfire that occurred in the Yeongnam region in March 2025 demonstrated that damage can be concentrated in specific areas where multiple risk factors such as weather, topography, vegetation, and human activities interact. However, wildfire occurrences exhibit strong spatial clustering, which can lead to overestimation of prediction performance and reduce map reliability due to information leakage caused by spatial autocorrelation when random partition-based validation is applied. To address this limitation, this study constructed a deep neural network-based wildfire susceptibility map at 60m resolution for Andong-si and Uiseong- gun in Gyeongsangbuk-do, and evaluated its spatial generalization performance using spatial cross-validation. The results indicated that, even under spatially separated validation conditions, the model maintained its discriminative power in distinguishing wildfire affected and unaffected areas. High susceptibility areas exhibited a clustering pattern centered around slopes, mountainous regions, and transition zones to lowlands. These high susceptibility patterns generally corresponded with NBR (Normalized Burn Ratio)-based damage distributions, whereas threshold based indicators showed limited performance due to severe class imbalance and strict spatial separation conditions. This study demonstrates the potential of susceptibility maps as evidence-based tools for wildfire prevention and resource allocation prioritization, emphasizing the necessity of incorporating spatial autocorrelation into validation designs for deep learning-based wildfire prediction.
Keywords
wildfire susceptibility
spatial autocorrelation
deep neural network
spatial cross-validation
NBR
2025년 3월 영남 지역에서 발생한 대형 산불은 기상, 지형, 식생, 인간 활동 등 여러 위험 요인이 복합적으로 작용하는 특정 공간에서 피해가 집중될 수 있음을 보여주었다. 따라서 산불 위험 지역을 사전에 식별하는 산불 민감성 지도의 구축과, 그 예측력에 대한 엄밀한 검증이 중요하다. 그러나 일반적인 교차검증은 공간적 자기상관에 따른 정보 누출로 예측 성능을 과대평가하고 지도의 신뢰성을 저하할 수 있다. 이에 본 연구는 경상북도 안동시와 의성군을 대상으로 60m 해상도의 심층신경망 기반 산불 민감성 지도를 구축하고 공간 교차검증을 적용하여 공간 일반화 성능을 평가하였다. 분석 결과, 공간적으로 분리된 검증 조건에서도 모델은 산불 피해·비 피해 지역을 구분하는 판별력을 유지했으며, 고민감 지역은 산지 능선, 사면, 산지와 저지대 전이 구간을 중심으로 군집하는 패턴을 보였다. 이러한 고위험 패턴은 NBR(Normalized Burn Ratio) 기반 피해 분포와 전반적으로 대응하였으나, 임계값 기반 지표는 심각한 클래스 불균형과 엄격한 공간적 분리 조건의 영향으로 제한적으로 나타났다. 본 연구는 산불 예방 및 자원 배분의 우선순위를 결정하는 증거 기반 도구로서의 민감성 지도의 활용 가능성을 제시하며, 딥러닝 기반 산불 예측에서 공간적 자기상관을 고려한 검증 설계의 필요성을 강조한다.
키워드
산불 민감성
공간적 자기상관
심층신경망
공간 교차검증
NBR
References
  1. 경북도청, 2025, 경북 초대형 산불, 초고속 회복을 위한 복구계획 최종 확정, 보도자료 2025년 5.7.
  2. 국립산림과학원, 2025, 2025년 산불 제대로 알기.
  3. 국회입법조사처, 2025, 산불 대응 및 지휘체계의 현황과 개선과제.
  4. 김삼근·안재근, 2020, 기상 데이터를 이용한 데이터 마이닝 기반의 산불 예측 모델. 한국산학기술학회 논문지, 21(8), 521-529.
  5. 김성훈·장동호, 2014, Landsat 위성영상을 이용한 충청남도임상 분석 및 산림 탄소저장량 추정, 한국지역지리학회지, 20(2), 206-216.
  6. 박노욱·장동호, 2011, 시계열 기온 분포도 작성을 위한 시공간 자기상관성 정보의 결합, 한국지역지리학회지, 17(4), 432-442.
  7. 심우진·박선영·정부매, 2024, 의성군의 하천 유로 변화와 취락 분포 양상, 한국지역지리학회지, 30(1), 97-111.10.26863/JKARG.2024.2.30.1.97
  8. 원명수·윤석희·구교상·김경하, 2011, 1990 년대와 2000 년대 건조계절의 산불발생 시공간 변화 분석, 한국지리정보학회지, 14(3), 150-162.10.11108/kagis.2011.14.3.150
  9. 조관연, 2011, 문화콘텐츠산업의 전략적 수용과 안동 문화정체성의 재구성, 한국지역지리학회지, 17(5), 568-581.
  10. 최진무·박선엽, 2020, 우리나라 계절적 강수와 건조도가 지역적 산불 발생에 미치는 영향, 한국지역지리학회지, 26(4), 307-319.10.26863/JKARG.2020.11.26.4.307
  11. 탁한명·김성환·손일, 2013, 지형학적 산지의 분포와 공간적 특성에 관한 연구, 대한지리학회지, 48(1), 1-18.10.6112/kscfe.2013.18.1.001
  12. Abatzoglou, J.T. and Williams, A.P., 2016, Impact of anthropogenic climate change on wildfire across western US forests, Proceedings of the National Academy of Sciences, 113(42), 11770-11775.10.1073/pnas.160717111327791053PMC5081637
  13. Anselin, L., 1995, Local indicators of spatial association—LISA, Geographical analysis, 27(2), 93-115.10.1111/j.1538-4632.1995.tb00338.x
  14. Arlot, S. and Celisse, A., 2010, A survey of cross-validation procedures for model selection, Statistics Surveys, 4, 40-7910.1214/09-SS054
  15. Bahadori, N., Razavi-Termeh, S.V., Sadeghi-Niaraki, A., Al-Kindi, K.M., Abuhmed, T., Nazeri, B., and Choi, S.M., 2023, Wildfire susceptibility mapping using deep learning algorithms in two satellite imagery dataset, Forests, 14(7), 1325.10.3390/f14071325
  16. Bengio, Y., 2009, Learning deep architectures for AI, Foundations and trends@ in Machine Learning, 2(1), 1-127.10.1561/2200000006
  17. Bhowmik, R.T., Jung, Y.S., Aguilera, J.A., Prunicki, M., and Nadeau, K., 2023, A multi-modal wildfire prediction and early-warning system based on a novel machine learning framework, Journal of environmental management, 341, 117908.10.1016/j.jenvman.2023.117908
  18. Bouzeraa, Y., Bouchemal, N., Djaaboub, S., Hristov, G., and Zahariev, P., 2025, Machine Learning-Based Wildfire Susceptibility Mapping: A GIS-Integrated Predictive Framework, Applied Sciences, 15(22), 12188.10.3390/app152212188
  19. Chuvieco, E., Mouillot, F., van der Werf, G.R., San Miguel, J., Tanase, M., Koutsias, N., García, M., Yebra, M., Padilla, M., Gitas, I., Heil, A., Hawbaker, T.J., and Giglio, L., 2019, Historical background and current developments for mapping burned area from satellite Earth observation, Remote Sensing of Environment, 225, 45–64.10.1016/j.rse.2019.02.013
  20. Cybenko, G., 1989, Approximation by superpositions of a sigmoidal function, Mathematics of control, signals and systems, 2(4), 303-314.10.1007/BF02551274
  21. Davies, D.K., Ilavajhala, S., Wong, M.M., and Justice, C.O., 2008, Fire information for resource management system: archiving and distributing MODIS active fire data, IEEE Transactions on Geoscience and Remote Sensing, 47(1), 72-79.10.1109/TGRS.2008.2002076
  22. Deeming, J.E., Burgan, R.E., and Cohen, J.D., 1977, The national fire-danger rating system, 1978 (Vol. 39), Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station.
  23. Durlević, U., Ilić, V., and Valjarević, A., 2025, Wildfire susceptibility mapping using deep learning and machine learning models based on multi-sensor satellite data fusion: a case study of Serbia, Fire, 8(10), 407.10.3390/fire8100407
  24. Elkan, C., 2001, The foundations of cost-sensitive learning, In International joint conference on artificial intelligence, 17(1), 973-978.
  25. Finney, M.A., 1998, FARSITE, Fire Area Simulator--model development and evaluation (No. 4), The Station.10.2737/RMRS-RP-4
  26. Goodfellow, I., Bengio, Y., and Courville, A., 2016, Deep learning (Vol. 1, No. 2), Cambridge: MIT press.
  27. Gregr, E.J., Palacios, D.M., Thompson, A., and Chan, K.M., 2019, Why less complexity produces better forecasts: an independent data evaluation of kelp habitat models, Ecography, 42(3), 428-443.10.1111/ecog.03470
  28. Hornik, K., Stinchcombe, M., and White, H., 1989, Multilayer feedforward networks are universal approximators, Neural networks, 2(5), 359-366.10.1016/0893-6080(89)90020-8
  29. Jain, P., Coogan, S.C., Subramanian, S.G., Crowley, M., Taylor, S., and Flannigan, M.D., 2020, A review of machine learning applications in wildfire science and management, Environmental Reviews, 28(4), 478-505.10.1139/er-2020-0019
  30. Jemeļjanova, M., Kmoch, A., and Uuemaa, E., 2024, Adapting machine learning for environmental spatial data-A review, Ecological Informatics, 81, 102634.10.1016/j.ecoinf.2024.102634
  31. Jo, H.W., Krasovskiy, A., Hong, M., Corning, S., Kim, W., Kraxner, F., and Lee, W.K., 2023, Modeling historical and future forest fires in south Korea: The flam optimization approach, Remote Sensing, 15(5), 1446.10.3390/rs15051446
  32. Kapoor, S. and Narayanan, A., 2023, Leakage and the reproducibility crisis in machine-learning-based science, Patterns, 4(9).10.1016/j.patter.2023.10080437720327PMC10499856
  33. Kattenborn, T., Schiefer, F., Frey, J., Feilhauer, H., Mahecha, M.D., and Dormann, C.F., 2022, Spatially autocorrelated training and validation samples inflate performance assessment of convolutional neural networks, ISPRS Open Journal of Photogrammetry and Remote Sensing, 5, 100018.10.1016/j.ophoto.2022.100018
  34. Key, C.H. and Benson, N.C., 2006, Landscape assessment (LA), in Lutes, D.C., Keane, R.E., Caratti, J.F., Key, C.H., Benson, N.C., Sutherland, S., and Gangi, L.J., FIREMON: Fire effects monitoring and inventory system, Gen. Tech. Rep. RMRS-GTR-164-CD. Fort Collins, CO: US Department of Agriculture, Forest Service, Rocky Mountain Research Station. 164, LA-1-55.
  35. Kim, J., Kim, T., Lee, Y.E., and Im, S., 2025, Spatial and temporal variability of forest fires in the Republic of Korea over 1991–2020, Natural Hazards, 1-21.10.1007/s11069-025-07169-4
  36. Kingma, D.P., 2014, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980.
  37. Koutsias, N., Kalabokidis, K.D., and Allgöwer, B., 2004, Fire occurrence patterns at landscape level: beyond positional accuracy of ignition points with kernel density estimation methods, Natural Resource Modeling, 17(4), 359-375.10.1111/j.1939-7445.2004.tb00141.x
  38. Lee, C., Choi, E.H., Han, Y., and Lee, Y., 2025, Year-round daily wildfire prediction and key factor analysis using machine learning: a case study of Gangwon State, South Korea, Scientific Reports, 15(1), 29910.10.1038/s41598-025-15508-540813792PMC12354747
  39. Legendre, P., 1993, Spatial autocorrelation: trouble or new paradigm?, Ecology, 74(6), 1659-1673.10.2307/1939924
  40. Leuenberger, M., Parente, J., Tonini, M., Pereira, M. G., and Kanevski, M., 2018, Wildfire susceptibility mapping: Deterministic vs. stochastic approaches, Environmental Modelling & Software, 101, 194-203.10.1016/j.envsoft.2017.12.019
  41. Martínez, J., Vega-Garcia, C., and Chuvieco, E., 2009, Human-caused wildfire risk rating for prevention planning in Spain, Journal of environmental management, 90(2), 1241-1252.10.1016/j.jenvman.2008.07.005
  42. Montgomery, D.C. and Voth, S.R., 1994, Multicollinearity and leverage in mixture experiments, Journal of Quality Technology, 26(2), 96-108.10.1080/00224065.1994.11979510
  43. Moran, P.A., 1950, Notes on continuous stochastic phenomena, Biometrika, 37(1/2), 17-23.10.1093/biomet/37.1-2.17
  44. Park, J., Suh, J., and Baek, M., 2025, Climatic and Forest Drivers of Wildfires in South Korea (1980-2024): Trends, Predictions, and the Role of the Wildland–Urban Interface, Forests, 16(9), 1476.10.3390/f16091476
  45. Radeloff, V.C., Hammer, R.B., Stewart, S.I., Fried, J.S., Holcomb, S.S., and McKeefry, J.F., 2005, The wildland–urban interface in the United States, Ecological applications, 15(3), 799-805.10.1890/04-1413
  46. Roberts, D.R., Bahn, V., Ciuti, S., Boyce, M.S., Elith, J., Guillera-Arroita, G., Hauenstein, S., Lahoz-Monfort, J.J., Schröder, B., Thuiller, W., Warton, D. I., Wintle, B.A., Hartig, F., and Dormann, C.F., 2017, Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure, Ecography, 40(8), 913-929.10.1111/ecog.02881
  47. Rothermel, R.C., 1972, A mathematical model for predicting fire spread in wildland fuels (Vol. 115), Intermountain Forest & Range Experiment Station, Forest Service, US Department of Agriculture.
  48. Saaty, T.L., 1987, Rank generation, preservation, and reversal in the analytic hierarchy decision process, Decision sciences, 18(2), 157-177.10.1111/j.1540-5915.1987.tb01514.x
  49. Salvati, L. and Ferrara, A., 2015, Validation of MEDALUS fire risk index using forest fires statistics through a multivariate approach, Ecological Indicators, 48, 365-369.10.1016/j.ecolind.2014.08.027
  50. Scott, J.H., Thompson, M.P., and Calkin, D.E., 2013, A wildfire risk assessment framework for land and resource management.10.2737/RMRS-GTR-315
  51. Seo, Y., Lee, D., and Choi, J., 2018, Growth pattern analysis of major coniferous tree species in South Korea, Forest Science and Technology, 14(1), 1-6.10.1080/21580103.2017.1409660
  52. Shrestha, N., 2020, Detecting multicollinearity in regression analysis, American journal of applied mathematics and statistics, 8(2), 39-42.10.12691/ajams-8-2-1
  53. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., and Salakhutdinov, R., 2014, Dropout: a simple way to prevent neural networks from overfitting, The journal of machine learning research, 15(1), 1929-1958.
  54. Stock, A., 2025, Choosing blocks for spatial cross-validation: lessons from a marine remote sensing case study, Frontiers in Remote Sensing, 6, 1531097.10.3389/frsen.2025.1531097
  55. Symeonidis, P., Vafeiadis, T., Ioannidis, D., and Tzovaras, D, 2025, Wildfire Susceptibility Mapping in Greece Using Ensemble Machine Learning, Earth, 6(3), 75.10.3390/earth6030075
  56. Syphard, A.D., Radeloff, V.C., Keeley, J.E., Hawbaker, T.J., Clayton, M.K., Stewart, S.I., and Hammer, R.B., 2007, Human influence on California fire regimes, Ecological applications, 17(5), 1388-1402.10.1890/06-1128.1
  57. Syphard, A.D., Rustigian-Romsos, H., Mann, M., Conlisk, E., Moritz, M.A., and Ackerly, D., 2019, The relative influence of climate and housing development on current and projected future fire patterns and structure loss across three California landscapes, Global Environmental Change, 56, 41-55.10.1016/j.gloenvcha.2019.03.007
  58. Taylor, S.W. and Alexander, M.E., 2006, Science, technology, and human factors in fire danger rating: the Canadian experience, International Journal of Wildland Fire, 15(1), 121-135.10.1071/WF05021
  59. Telford, R.J. and Birks, H.J.B., 2009, Evaluation of transfer functions in spatially structured environments, Quaternary Science Reviews, 28(13-14), 1309-1316.10.1016/j.quascirev.2008.12.020
  60. Tobler, W.R., 1970, A computer movie simulating urban growth in the Detroit region, Economic geography, 46(sup1), 234-240.10.2307/143141
  61. Tonini, M., D’Andrea, M., Biondi, G., Degli Esposti, S., Trucchia, A., and Fiorucci, P., 2020, A machine learning-based approach for wildfire susceptibility mapping. The case study of the Liguria region in Italy, Geosciences, 10(3), 105.10.3390/geosciences10030105
  62. Valavi, R., Elith, J., Lahoz-Monfort, J.J., and Guillera-Arroita, G., 2018, blockCV: An r package for generating spatially or environmentally separated folds for k-fold cross-validation of species distribution models, Biorxiv, 357798.10.1101/357798
  63. Wagner, C.V., 1987, Development and structure of the Canadian forest fire weather index system, 35, viii+-37.
  64. 공공데이터포털, https://www.data.go.kr
  65. 국토지리정보원 국토정보플랫폼, https://map.ngii.go.kr
  66. 기상청 기상자료개방포털, https://data.kma.go.kr
  67. 미국 지질조사국, https://www.usgs.gov
  68. 브이월드, https://www.vworld.kr
  69. 산림공간정보서비스, https://www.forest.go.kr
  70. 통계지리정보서비스, https://sgis.mods.go.kr
  71. 환경공간정보서비스, https://aid.mcee.go.kr
Information
  • Publisher :The Korean Association of Regional Geographers
  • Publisher(Ko) :한국지역지리학회
  • Journal Title :REGION AND GEOGRAPHY
  • Journal Title(Ko) :지역과 지리
  • Volume : 32
  • No :1
  • Pages :81-105
  • DOI :https://doi.org/10.26863/JKARG.2026.2.32.1.81
Journal Informaiton REGION AND GEOGRAPHY REGION AND GEOGRAPHY
Journal Informaiton Journal Informaiton - close