شناسایی متغیرهای تأثیرگذار بر آتش‌سوزی در اراضی طبیعی شهرستان خرم‌آباد با استفاده از مدل درخت تصمیم

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشجوی دکتری علوم و مهندسی جنگل، دانشکده منابع طبیعی، دانشگاه لرستان، خرم‌آباد، ایران

2 گروه مهندسی جنگل‌داری، دانشکده منابع طبیعی، دانشگاه لرستان، خرم‌آباد

3 استاد بخش مهندسی منابع طبیعی و محیط زیست، دانشکده کشاورزی، دانشگاه شیراز، شیراز، ایران

4 گروه فتوگرامتری و سنجش از دور، دانشکده مهندسی نقشه‌برداری، دانشگاه صنعتی خواجه نصیرالدین طوسی، تهران، ایران

‎10.22052/deej.2025.256094.1090

چکیده

وقوع مکرر آتش‌سوزی در اراضی طبیعی منطقۀ زاگرس، ضرورت انجام پژوهش‌هایی برای شناسایی عوامل مؤثر و پیش‌بینی آتش‌سوزی‌های آتی را اجتناب‌ناپذیر می‌کند. لذا پژوهش حاضر با هدف شناسایی عوامل مؤثر بر وقوع آتش‌سوزی در حوزۀ آبخیز شهرستان خرم‌آباد‌ و با استفاده از روش‌ مدل‌سازی مبتنی‌بر درخت تصمیم، انجام گرفت. بدین منظور تأثیر عوامل اقلیمی، توپوگرافی، کاربری و پوشش زمین و انسانی بر وقوع آتش‌سوزی در منطقه مورد بررسی قرار گرفت. ازطرف دیگر، تعداد 380 نقطۀ آتش‌سوزی مربوط به بازۀ زمانی 1390 تا 1403 که از اداره‌کل منابع طبیعی و آبخیزداری استان لرستان اخذ شده بود، به‌عنوان متغیر وابسته در روند مدل‌سازی به کار گرفته شد. 70 درصد از این نقاط (266 نقطه) به‌عنوان داده‌های آموزشی و 30 درصد (114 نقطه) برای داده‌های آزمون در نظر گرفته شد. برای ارزیابی و صحت‌سنجی مدل از منحنی ROC و ماتریس ابهام استفاده شد. تحلیل اهمیت متغیرها نشان داد که متغیرهای فاصله از جاده و شاخص نرمال‌شدۀ تفاضل پوشش گیاهی (NDVI) از مهم‌ترین عوامل مؤثر بر وقوع آتش‌سوزی در منطقه است. نتایج حاکی از آن بود که با افزایش میزان فاصله از جاده‌ها، افزایش تراکم پوشش گیاهی و نیز افزایش میزان تأثیر باد، میزان وقوع آتش‌سوزی نیز افزایش خواهد یافت. همچنین نتایج حاصل از صحت‌سنجی مدل نشان داد که توانایی مدل مذکور در تشخیص صحیح وقوع آتش‌سوزی در مقابل عدم وقوع آن، برابر با 85 درصد است. نتایج این پژوهش، می‌تواند برای شناسایی عوامل مؤثر بر آتش‌سوزی و پیش‌گیری از آتش‌سوزی‌های آینده و مدیریت پایدار اراضی مورد استفاده قرار گیرد.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

Identifying Factors Influencing Wildfires in Khorramabad's Natural Areas Using a Decision Tree Model

نویسندگان [English]

  • soheila naserirad 1
  • Hamed Naghavi 2
  • Hamidreza Pourghasemi 3
  • Arvin Fakhri 4
1 PhD student in Forest Science and Engineering, Faculty of Natural Resources, Lorestan University, Khorramabad, Iran
2 Department of Forestry Engineering, Faculty of Natural Resources, Lorestan University, Khorramabad, Iran
3 Professor, Department of Engineering and Environment, Faculty of Agriculture, Shiraz University, Shiraz, Iran
4 Department of Photogrammetry and Remote Sensing, Faculty of Geodesy and Geomatics Engineering, K. N Toosi University of Technology, Tehran, Iran
چکیده [English]

Itroduction: The frequent occurrence of wildfires in the natural landscapes of the Zagros region necessitates comprehensive research to identify contributing factors and predict future fire incidents. Accordingly, this study aims to determine the key drivers of wildfires in the Khorramabad County watershed, located in the central Zagros, using decision tree-based modeling approaches.
 
Material and Methods: The study examined fire-influencing factors across four categories: climate, topography, land use, and human activity. Wildfire data spanning 2011–2024 (380 fire points), obtained from the Lorestan Province General Department of Natural Resources and Watershed Management, served as the dependent variable. The dataset was split into training (70%, 266 points) and evaluation (30%, 114 points) subsets. Model performance was assessed using the ROC curve and confusion matrix for validation..
 
Results: Variable importance analysis revealed that distance from roads and the Normalized Difference Vegetation Index (NDVI) were among the most significant factors influencing wildfire occurrences in the study area. The results demonstrated a positive correlation between fire incidence and three key variables: (1) greater proximity to roads, (2) higher vegetation density, and (3) increased wind effects. Model validation indicated an 85% accuracy rate in correctly classifying fire events versus non-fire events. These findings provide actionable insights for identifying fire risk factors, supporting wildfire prevention strategies, and promoting sustainable land management practices in the region..
 
Discusssion and Conclusion: Given the critical role of forests and rangelands in water and soil conservation, as well as erosion prevention, identifying key drivers of wildfires is essential for effective land management. This study highlights distance from roads, NDVI, precipitation, wind effect, and temperature as the most influential factors affecting fire occurrences in the study area, ranked in order of significance. The decision tree model demonstrated high predictive accuracy (85%), confirming its effectiveness in identifying wildfire risk factors. These findings provide valuable insights for:Wildfire prevention strategies, Sustainable land management, and Informed decision-making for policymakers and land-use planners. Furthermore, the results serve as a foundational reference for future research on natural resource management, particularly in fire-prone ecosystems. By integrating these findings into regional planning, stakeholders can better mitigate fire risks and enhance ecosystem resilience.

کلیدواژه‌ها [English]

  • Machine learning
  • Remote sensing
  • NDVI
  • ROC curve

مراجع

  1. Achu, A.L., & Reghunath, R. )2021(. Machine learning modeling of fire susceptibility in a forest agriculture mosaic landscape of southern India. ecological information, 12-20.
  2. Anjum, N., Qammer, H., & Masood, R. )2022(. Intelligent computing based forecasting of deforestation using fire alerts. A deep learning approach 55, 2-6.
  3. Bagherabadi, R., Sheikh Katlo Milan, F., & Zarei Mohammad Abad, M. (2022). Fire risk assessment in Zagros forests. Ecosystem Management Journal 1 (3), 60-72.
  4. Bashari, H., A.A., Naghipour., S.J., Khajeddin., H., & Sangoony Tahmasebi, P. ( 2016). Risk of fire occurrence in arid and semi-arid ecosystems of Iran: an investigation using Bayesian belief networks. Environmental monitoring and assessment 9, 531.
  5. Can Ilban, M., & Sekertekin, A. (2022). Machine learning based wildfire susceptibility mapping using remotely sensed fire data and GIS case study of Adana and Mersin provinces Turkey. 1-24.
  6. Chavan, M., Das, K.K., & Suryawanshi, R. (2012). Forest fire risk zonation using remote sensing and GIS in Huynial watershed Tehri Garhwal district. International Journal of Basic and Applied Research 2 (7), 6-12.
  7. Cheraghi, R., Sharifi, J., Mohammadi, H., (2020). The effect of fire on the density and composition of vegetation in forest lands of the Vanan Khalkhal region. Journal of Nature of Iran, 6 (6),76-87.
  8. Dehghani, A.R. (2020). Preparing a forest fire hazard distribution map using multi-criteria decision-making method. 1-43.
  9. Dearojou, G., Brun, C., Margalef, T., & Cortes, A. (2015). Wind field uncertainty in forest fire propagation prediction. Procedia Computer Science 29, 1535–1545.
  10. Eskandari, S. (2015). Evaluation of Forest Fire Risk Potential. Geographical Spatial Planning Journal, 5, 17, 192-205.
  11. Eskandari, S., Pourghasemi, H.R., & Tiefenbacher, J.P. (2021). Fire susceptibility mapping in the Northeast forests and rangeland of Iran using new and ensemble data mining models. 1-22.
  12. Eslami, R., Azarnoosh, M.R., Kialashaki, A., & Kzem Nezhad, F. (2019). Modeling the spread of forest fires based on GIS and cellular automation. Case Study: Babolrud Watershed, Mazandaran Journal of Geographical Sciences 32, 107-117.
  13. Erten, E., Kurgun, V., & Musaoglue, N. (2004). Forest fire risk zone mapping from satellite imagery and GIS: a case study. XXth Congress of the International Society for Photogrammetry and Remote Sensing, Istanbul, Turkey. 222-230.
  14. Eyvazi, Sh., Babaei Kafaki, S., Jafari, M., & Tahmooreth, M. (2021). The status and value of renewable natural resources and their protection and exploitation. Journal of Natural Resources 34 (3), 599-611.
  15. Faezizadeh, B. (2021). Comparative evaluation of pixel-based and object-oriented processing techniques in ASTER satellite image classification for extracting agricultural and garden land maps on the eastern shore of Lake Urmia. Journal of Scientific research of geographic information 28 (109), 167-183.
  16. Feghhi, J., Ali Mahmoodi Sarab, S., & Gholi Khageh, S. (2019). Preparing a forest fire hazard map using artificial neural networks in Golestan province. Journal of Wood and Forest Science and Technology Research 25 (2), 123-136.
  17. Gholami, S., & Kafash Charandani, N., (2018). The width of the dam, (or fire, danger, burning, model, and construction of the water reservoirs of the forests of the Yarsbaran. Conference on Technology, Applications and Innovation in Geomatics, Tabriz.
  18. Goodfellow, I., Bengio, Y., & Courville, A. (2021). Deep learning. MIT press.3 (5), 42-56.
  19. Hong, H., Naghibi, S.A., Moradi Dashtpagerdi, M.M., Pourghasemi, H.R., & Chen, W. (2017). A comparative assessment between linear and quadratic discriminant analyses (LDA-QDA) with frequency ratio and weights-of-evidence models for forest fire susceptibility mapping in China. Arabian Journal of Geosciences, 10 (7), 1-14.
  20. Jahdi, R. (2023). Modeling fire risk and pressure using geographic information systems in Khalkhal and Kosar counties. Journal of Spatial Analysis of Environmental Hazards 3, 79-9.
  21. Jaiswal, R.K., Saumitra, M., Kumaran, R.D., & Rajesh, S. (2023). Forest fire risk zone mapping from satellite imagery and GIS. International Journal of Applied Earth Observation and Geoinformation 4, 1–10.
  22. Khalaj, A., & Pourghasem, M. (2017). The role of weather and atmospheric factors in forest fires. The First International Conference on the Position of Safety, Health and Environment in Organizations.
  23. Kuklina, V., Sizov, O., Bogdanov, V., Krasnoshtanova, N., Morozova, A., & Petrov, N. (2023). Combining community observations and remote sensing to examine the effects of roads on wildfires in the East Siberian boreal forest 9, 293-407.
  24. Kumari, B., & Pandey, A.C. (2020). Geo-informatics based multi-criteria decision analysis (MCDA) through analytic hierarchy process (AHP) for forest fire risk mapping in Palamau Tiger Reserve Jharkhand stateIndia. Journal of Earth System Science 129 (1), 1-16.
  25. Matthew, W., Douglas, I., & Chantele, A. (2024). State of Wildfires 2023-2024. Journal of Spatial Analysis of Environmental Hazards 17-26.
  26. Mirzaeezadeh, V., & Niknezhad, M. (2014). Identifying factors affecting forest cover reduction using Landsat satellite images. Journal of Conservation and Exploitation of Natural Resources 2, 91-108.
  27. Mohamadi, F., Shaabanian, M., Poorhashmi, M., & Fatehi, P. (2020). Forest fire risk mapping using GIS and AHP in DA part of the forest. Kordestan Quarterly Iran. 18 (4), 569-586.
  28. Mohammadian, M. (2010). Research on Zagros anticlines in the Seymourh watershed in Darreshahr county, Bachelor's thesis. Islamic Azad
  29. Morvati, M., & Karami, P. (2024). Analysis of spatial-temporal changes in fire occurrence in Kermanshah province with emphasis on the network of protected areas. Journal of Natural Environmental Hazards 13, 40.
  30. Mosa Beigi, M., & Mirza Beigi, F. (2016). Fire hazard analysis model using GIS network. Scientific knowledge 4 (14), 175-188.
  31. Najafi, A., Iran Nezhad, M.H., Sotoudeh, A., Mokhtari, M.H., & Kiani, B. (2017). Modeling and mapping forest fire risk using GIS-based measurements. Applied Geology 14, 13-25.
  32. Narayanaraj, G., & Wimberly, C. (2012). Influences of forest roads on the spatial patterns of human- and lightning-caused wildfire ignitions. Apllied Geography, 32, 787-888.
  33. Nourizadeh, M., Naghavi, H., & Omidvar, E. (2023). The effect of land use and land cover changes on soil erosion in semi‑arid areas using cloud‑based google earth engine platform and GIS‑based RUSLE model.
  34. Omidvar, K., Shafiei, Sh., Taghizadeh, Z., & Alipoor, M. (2018). Evaluation of the efficiency of the decision tree model in rainfall forecasting at the Kermanshah Synoptic Station. Journal of Applied Research in Geographic Sciences 14 (4), 89-110.
  35. Pazhouh, F., & Alijani, B. (2024). Analysis of forest fires in Western Guilan County for risk management. 3, 5-20.
  36. Pourtaghi, Z.S., Pourghasemi, H.R., & Rossi, M. (2015). Forest fire susceptibility mapping in the Minudasht forests Golestan province, Environmental Earth Sciences, 73 (4), 1515-1533.
  37. Pourghasemi, H.R., Beheshtirad, M., & Pradhan, B. (2014). A comparative assessment of prediction capabilities of modified analytical hierarchy process (M-AHP) and Mamdani fuzzy logic models using Netcad-GIS for forest fire susceptibility mapping. Geomatics, Natural Hazards and Risk 7 (2), 861-885.
  38. Rawat, G.S. (2003). Fire risk assessment for forest fire control management in Chilla Forest Range of Rajaji National Park, Uttaranchal, (India). M.Sc Thesis, International institute for Geo- Information Sciience and Earth Observation Enschede, ITC University, Netherlands 78.
  39. Romero-Calcerrada, R. C., Novillo, J., & Millington, I. (2008). GIS analysis of spatial patterns of human-caused wildfire ignition risk in the SW of Madrid (Central Spain). Landscape Ecology 23 (3), 341-356.
  40. Sagar, P.,Suresh, Y.B., Naveesh, C.A., Archana, D., Hemadri, S.S., Patil, V.P., Archana, R., & Raaga, A.S.(2024). Forest fire dynamics in India unveiling climate impact spatial pattern and interface with antrrax incidence. Ecological Indicators 166.
  41. Sarkargar, A., Valdan, M., & Mansourian, A. (2018). Spatial analysis of fire using RS, GIS in different parts of the country. Journal of Ecology 35, 25- 34.
  42. Somashekar, R., Ravikumar, P., Mohankumar, C., Prakash, K., & Nagaraja, B. (2009). Burntarea mapping of bandipur national park, India using IRS1C\ 1D LISS III data. Indiansoc. Remote sensing 37, 37- 50.
  43. Stocks, M.R., Lynham, A.F., Lawson, B.D., & Alexander, K.L. (2020). The Danger of forest fire. the forestry chronicle 54, 6, 243-278.
  44. Uthappa, A.R., Das, B., Raizada, A., Kumar, P., Jha, P. & Prasad, P.V. (2025). Forest fire susceptibility mapping using multi-criteria decision making and machine learning models in the Western Ghats of India. Journal of Environmental Management, 379, p.124777.
  45. Visa, S., Ramsay, B., & Ralescu, A. (2016). Confusion matrix based feature selection. 3, 32-41.
  46. Yang, J., Shifley S.R., & Gustafson, E.J. (2017). Spatial patterns of modern period human-caused fire occurrence in the Missouri Ozark Highlands. Forest science 53 (1), 1-15.
  47. Zumbrunnen, T., Pezzatti, G., Menéndez, P., Bugmann, H., Bürg, I.M., & Conedera, M. (2020). Weather and human impacts on forest fires: 100 years of fire history in two climaticregions of Switzerland. Forest Ecology and Management 261 (12), 2188-2199.
مهندسی اکوسیستم بیابان
دوره 14، شماره 46
مرداد 1404
صفحه 1-16

فایل ها

هم رسانی

ارجاع به این مقاله

آمار