환경영향평가 (Journal of Environmental Impact Assessment)

한국환경영향평가학회 (Korean Society of Environmental Impact Assessment)
권호 표지
DOI QR코드

DOI QR Code

기계학습 기반의 로드킬 발생 예측과 영향 요인 탐색에 대한 연구

A Study on Machine Learning-Based Estimation of Roadkill Incidents and Exploration of Influencing Factors

  • 투고 : 2024.02.15
  • 심사 : 2024.03.04
  • 발행 : 2024.04.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

본 연구에서는 충청남도를 중심으로 로드킬 발생을 예측하고 영향을 미치는 요인을 탐구하여 로드킬 예방 대책 수립에 이바지하고자 하였다. 날씨, 도로 및 환경 정보를 종합적으로 고려하여 기계학습을 기반으로 로드킬 발생을 예측하고 각 변수의 중요성을 분석하여 주요 영향 요인을 도출하였다. 가장 우수한 성능을 보인 Gradient Boosting Machine(GBM)은 정확도 92.0%, 재현율 84.6%, F1-score 89.2%, AUC 0.907을 기록했다. 로드킬에 영향을 미치는 주요 요인은 평균 지역 기압(hPa), 평균 지면 온도(℃), 월, 평균 이슬점 온도(℃), 중앙 분리대 존재 여부, 평균 풍속(m/s)이었다. 이러한 결과는 로드킬 예방 및 교통안전에 이바지할 것으로 기대되며, 생태계와 도로 개발 간의 균형 유지에 중요한 역할을 할 것으로 예상한다.

This study aims to estimate roadkill occurrences and investigate influential factors in Chungcheongnam-do, contributing to the establishment of roadkill prevention measures. By comprehensively considering weather, road, and environmental information, machine learning was utilized to estimate roadkill incidents and analyze the importance of each variable, deriving primary influencing factors. The Gradient Boosting Machine (GBM) exhibited the best performance, achieving an accuracy of 92.0%, a recall of 84.6%, an F1-score of 89.2%, and an AUC of 0.907. The key factors affecting roadkill included average local atmospheric pressure (hPa), average ground temperature (℃), month, average dew point temperature (℃), presence of median barriers, and average wind speed (m/s). These findings are anticipated to contribute to roadkill prevention strategies and enhance traffic safety, playing a crucial role in maintaining a balance between ecosystems and road development.

키워드

과제정보

본 연구는 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행하였습니다(RS-2023-00242528).

참고문헌

  1. Choi TY, Park JH. 2006. The Effects of Land Use on the Frequency of Mammal Roadkills in Korea. Journal of the Korean Institute of Landscape Architecture, 34(5): 52-58. [Korean
  2. Do MS, Yoo JC. 2014. Distribution Pattern According to Altitude and Habitat Type of the Red-tongue Viper Snake (Gloydius Ussuriensis) in the Cheonma Mountain. Journal of Wetlands Research, 16(2): 193-204. [Korean Literature]  https://doi.org/10.17663/JWR.2014.16.2.193
  3. Ha H, Shilling F. 2018. Modelling potential wildlife-vehicle collisions (WVC) locations using environmental factors and human population density: A case-study from 3 state highways in Central California. Ecological Informatics, 43: 212-221.  https://doi.org/10.1016/j.ecoinf.2017.10.005
  4. Kim M, Park H, Lee S. 2021. Analysis of Roadkill on the Korean Expressways from 2004 to 2019. International Journal of Environmental Research and Public Health, 18(19): 10252. 
  5. Kweon HK, Choi YH, Kim MJ, Lee JW. 2008. Study on the Status and Cause of the Road Kill for Wildlife Killing Reduce - A Case Study of National Road in Daejeon~Seosan Section. Journal of Forest and Environmental Science, 24(2): 99-109. [Korean Literature] 
  6. Lee GJ, Tak JH, Park SI. 2014. Spatial and Temporal Patterns on Wildlife Road-kills on Highway in Korea. Journal of Veterinary Clinics, 31(4): 282-287. [Korean Literature]  https://doi.org/10.17555/ksvc.2014.08.31.4.282
  7. Lundberg SM, Lee SI. 2017. A unified approach to interpreting model predictions. Advances in neural information processing systems, p. 30. 
  8. Min JH, Han GS. 2010. A Study on the Characteristics of Road-kills in the Odaesan National Park. Korean Journal of Environment and Ecology, 24(1): 46-53. [Korean Literature] 
  9. Ministry of Land, Infrastructure and Transport. 2022. Road Status Report 2022. General Statistics No. 116006. [Korean Literature] 
  10. National Institute of Ecology. 2022. Intensive Survey on Road-kill Hotspots in South Korea. [Korean Literature] 
  11. Son SW, Kil SH, Yoon YJ, Yoon JH, Jeon HJ, Son YH, Kim MS. 2016. Analysis of Influential Factors of Roadkill Occurrence - A Case Study of Seorak National Park. Journal of the Korean Institute of Landscape Architecture, 44(3): 1-12. [Korean Literature]  https://doi.org/10.9715/KILA.2016.44.3.001
  12. Song EG, Seo HJ, Kim KM, Woo DG, Park TJ, Choi TY. 2019. Analysis of Roadkill Hotspot According to the Spatial Clustering Methods. Journal of Environmental Impact Assessment, 28(6): 580-591. [Korean Literature] 
  13. Song JS, Lee KJ, Ki KS, Jun IY. 2011. The Efficiency and Improvement of the Highway Wild-Life Fences for Decrease of Mammals Road-kill - In Case of Manjong~Hongchun Section on Jungang Highway -. Korean Journal of Environment and Ecology, 25(5): 649-657. [Korean Literature] 
  14. Pagany R, Dorner W. 2019. Do crash barriers and fences have an impact on wildlife-vehicle collisions? - An artificial intelligence and GIS-based analysis. ISPRS International Journal of Geo-Information, 8(2): 66. 
  15. Pagany R, Valdes J, Dorner W. 2020. Risk Prediction of Wildlife-vehicle Collisions Comparing Machine Learning Methods and Data Use. In 2020 10th International Conference on Advanced Computer Information Technologies (ACIT) 436-440. IEEE.