مقایسه کارایی روش های یادگیری ماشینی درمدلسازی مناطق حساس به وقوع آتش (استان ایلام، شهرستان دره شهر)
محورهای موضوعی : منابع طبیعی و مدیریت زیست محیطی
مریم محمدیان
1
,
مریم مروتی
2
,
رضا امیدی پور
3
1 - دانشجوی کارشناسی ارشد علوم و مهندسی محیط زیست، دانشکده کشاورزی و منابع طبیعی، دانشگاه اردکان، ایران
2 - دانشیار گروه علوم و مهندسی محیط زیست، دانشکده کشاورزی و منابع طبیعی، دانشگاه اردکان، اردکان، ایران
3 - استادیار گروه مرتع و آبخیزداری، دانشکده کشاورزی، دانشگاه ایلام، ایلام، ایران
کلید واژه: استان ایلام, مدیریت اکوسیستم, ماشین بردار پشتیبان, جنگل تصادفی, مدلسازی آتش,
چکیده مقاله :
آتش از مهمترین مخاطرات طبیعی بوده که تاثیر بسزائی بر ساختار و پویایی اکوسیستم‌های طبیعی دارد. با توجه به قرارگیری ایران در کمربند خشک و نیمه‌خشک جهان، هر ساله تعداد زیادی آتشسوزی عمدی و غیر عمدی در مناطق مختلف کشور به وقوع می‌پیوندد. به همین دلیل تعیین مناطق حساس به وقوع آتش نقش مهمی در مدیریت آتشسوزی در منابع طبیعی دارد. به همین دلیل تحقیق حاضر در صدد است تا با استفاده از دو روش یادگیری ماشینی جنگل تصادفی (Random Forest) و ماشین بردار پشتیبان (Support Vector Machine) و 2024 نقطه وقوع آتش، مناطق حساس به وقوع آتش را در شهرستان دره شهر در استان ایلام را تعیین نماید. عوامل محیطی در چهار گروه اصلی شامل عوامل توپوگرافی (ارتفاع، جهت شیب، تند شیب)، عوامل اقلیمی (بارش، رطوبت نسبی، باد، درجه حرارت)، عوامل زیستی (پوشش گیاهی و رطوبت خاک) و عوامل انسان ساخت (فاصله از مناطق مسکونی، فاصله از جاده، فاصله از اراضی کشاورزی، فاصله از آبراهه) تهیه شدند. ارزیابی دقت مدل‌های مورد استفاده با استفاده سطح زیر نمودار (AUC) در منحنی ROC و آماره‌های ارزیابی متقاطع (Cross-validation) انجام شد. بررسی شاخص AUC نشان داد که هر دو مدل دارای دقت مناسبی بوده هرچند مدل جنگل تصادفی (AUC = 0.97) دارای دقت بالاتری نسبت به مدل ماشین بردار پشتیبان (AUC = 0.86) بود. بر اساس نتایج مدل جنگل تصادفی، حدود 60 درصد در کلاس کم خطر و حدود 20 درصد در کلاس خطر زیاد آتش قرار دارد. بررسی سهم عوامل تاثیرگذار بر وقوع آتش نشان داد که عوامل انسان ساخت (فاصله از مناطق مسکونی) و عوامل اقلیمی (درجه حرارت) نقش مهمتری در نقاط دارای سابقه آتش داشتند. بنابراین افزایش فرهنگ عمومی و کاهش رفتارهای خطرناک در طبیعت می‌تواند موجب کاهش وقوع آتش در این منطقه شده و سهم زیادی در حفاظت از محیط زیست و حفظ منابع طبیعی داشته باشد.
Fire is one of the most important natural hazards that has a great impact on the structure and dynamics of natural ecosystems. Due to Iran's location in the arid and semi-arid belt of the world, a large number of human-made and natural fires occur in different regions of the country every year. Therefore, determining sensitive areas to fire occurrence plays an important role in fire management in natural resources. To do so, the current study aims to identify fire-prone areas in Dere Shahr city in Ilam province using two machine learning of random forest (RF) and support vector machine (SVM) and 2024 fire occurrence points. Environmental factors were prepared in categories including topographical factors (altitude, slope direction, slope anlgle), climatic factors (rainfall, relative humidity, wind, temperature), biological factors (vegetation and soil moisture) and man-made factors (distance from residential areas, distance from road, distance from agricultural land, distance from river). The model’s accuracy was evaluated using the area under the curve (AUC) in the ROC curve and cross-validation statistics. Examining the AUC index showed that both models had good accuracy, although the RF model (AUC = 0.97) had higher accuracy than the support vector machine model (AUC = 0.86). According to the results of RF model, about 60% are in the low-risk class and about 20% are in the high fire risk class. Investigating the contribution of the factors affecting the occurrence of fire showed that man-made factors (distance from residential areas) and climatic factors (temperature) played a more important role in areas with a history of fire. Therefore, increasing public culture and reducing dangerous behaviors in nature can reduce the occurrence of fire in this area and contribute greatly to the protection of the environment and preservation of natural resources.
1. Adab H, Kanniah K.D, Solaimani K. 2013. Modeling forest fire risk in the northeast of Iran using remote sensing and GIS techniques. Natural Hazards, 65: 1723- 1743. Doi: https://doi.org/10.1007/s11069-012-0450-8.
2. Bazyar M, Oladi Ghadikolaii J, Pourghasemi H.R, Serajyan Maralan M.R. 2020. Zoning and Investigation of Factors Affecting Forest Fire Using Evidential Belief Function Algorithm and Support Vector Machine in Boyer Ahmad City. Iranian Journal of Forest and Range Protection Research, 17 (2): 197- 222. (In Persian).
3. Carnicer J, Vives‐Ingla M, Blanquer L, Méndez‐Camps X, Rosell C, Sabaté S, Gutiérrez E, Sauras T, Peñuelas J, Barbeta A. 2021. Forest resilience to global warming is strongly modulated by local‐scale topographic, microclimatic and biotic conditions. Journal of Ecology, 109(9):3322-39.
4. Conciani DE, dos Santos LP, Silva TS, Durigan G, Alvarado ST. Human-climate interactions shape fire regimes in the Cerrado of São Paulo state, Brazil. Journal for Nature Conservation. 2021 Jun 1;61:126006.
5. Emami H, Shahriyari H. 2020. Quantifying environmental and human factors affecting occurance and spread of wildfires using RS and GIS methods protected area of Arasbaran. Scientific - Research Quarterly of
Geographical Data (SEPEHR), 28 (112): 35- 53. (In Persian).
6. Erten E, Kurgun V, Musaolu N. 2005. Forest fire risk zone mapping from satellite imagery and firerisk: A recent case study from Mt. Carmel (Israel). Forest Ecology and Management, 262: 1184– 1188.
7. Eshaghi M.A, Shetaeejoybari, Sh. 2017. Preparation map of Forest Fire Risk Using SVM, RF & MLP Algorithms (Case Study: Golestan National Park, Northeastern Iran). Wood & Forest Science and Technology, 23 (4): 133- 154. (In Persian).
8. Eskandari S, Jalilvand H. 2012. Climate change and its effect on the frequency and area of forest fires in eastern Mazandaran. Proceedings of the third international conference on non-climate and tree chronology in Caspian ecosystems, Sari, May 2012. (In Persian).
9. Eskandari S, Oladi Ghadikolaei J, Jalilvand H, Saradjian M.R. 2013. Fire risk modeling and prediction in district three of Neka-Zalemroud forest, using Geographical Information System. Iranian Journal of Forest and Poplar Research, 21 (2): 203- 217. (In Persian).
10. Esmaeili Sharif M, Jalilvand H, Amoozad M, Jafari A, Sayed Mahale M. 2018. The effect of ecological factors on fire in Hyrcanian forests (Case study: forest areas of Neka, Mazandaran, Iran). Journal of Forest Research and Development, 4 (1): 113- 129. (In Persian).
11. Feizizadeh B, Omarzadeh D, Mohammadnejad V, Khallaghi H, Sharifi A, Karkarg BG. 2022. An integrated approach of artificial intelligence and geoinformation techniques applied to forest fire risk modeling in Gachsaran, Iran. Journal of Environmental Planning and Management, 8:1-23.
12. Gao B.C. 1996. NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space. Remote sensing of environment, 58(3): 257-266.
13. Guo F, Wang G.Su.Z, Liang H, Wang, W, Lin F.E.A. 2016. What drives forest fire in Fujian, China? Evidence from logistic regression and Random Forests. International Journal of Wildland Fire, 25 (5): 505- 519. DOI:10.1071/WF15121
14. Hernandez-Leal P, Arbelo M, Gonzalez-Calvo A. 2006. Fire risk assessment using satellite data. Advances in Space research, 37 (4): 741- 746. Doi: https://doi.org/10.1016/j.asr.2004.12.053
15. Holden Z, Morgan P, Evans J. 2009. A predictive model of burn severity based on 20-year satellite-inferred burn severity data in a large southwestern US wilderness area. For Ecol Manag, 258 (11): 2399– 2406. DOI:10.1016/j.foreco.2009.08.017
16. Hong H, Naghibi S.A, Moradi Dashtpagerdi 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: 1- 14. Doi: https://doi.org/10.1007/s12517-017-2905-4
17. Jaiswal R, Mukherjee S, Kumaran R, Saxena R, 2002. Forest fire risk zone mapping from satellite imagery and GIS. International Journal of Applied Earth Observation and Geoinformation, 4 (1): 1- 10. Doi: https://doi.org/10.1016/S0303-2434(02)00006-5
18. Janbazghobadi Gh. 2019. Investigation of forest fire hazard areas in Golestan province based on fire risk system index (FRSI) using the technique (GIS). Journal of Spatial Analysis Environmental Hazarts, 6 (3): 89- 102. (In Persian).
19. Kaboodi A, Azizi Y, Rezaee S. 2017. Investigating the main causes of Zagros forest fires and providing solutions to manage these forests (case study: Ilam province). The third national congress of development and promotion of agricultural engineering and soil science of Iran: 1- 14. (In Persian).
20. Kargar M, Farzi H, Taheri A. 2018. Comparison of random forest models and artificial neural network in predicting the occurrence of pasture fires in Alborz province. The 7th National Conference on Pasture and Pasture Management of Iran: 1- 13. (In Persian).
21. Karimi A, Abdollahi S, Ostad-Ali-Askari K, Eslamian S, Singh VP. 2021. Predicting fire hazard areas using vegetation indexes, case study: Forests of Golestan Province, Iran. Journal of Geography and Cartography, 5;4(1):1-6.
22. Leuenberger M, Parente J, Tonini M, Pereira M.G, Kanevski M. 2018. Wildfire susceptibility mapping: Deterministic vs. stochastic approaches. Environmental Modelling & Software, 101 (2018): 194- 203. Doi: https://doi.org/10.1016/j.envsoft.2017.12.019
23. Mafi-Gholami D, Jaafari A. 2020. Analysis of relationship between the meteorological and hydrological droughts and biomass change of mangrove forests in the Gwatr Port over a 34-year period. Iranian Journal of Forest and Poplar Research, 28 (2): 149- 162. (In Persian).
24. Mahdavi A, Rangin S, Mehdizadeh H, Mirzaei Zadeh V. 2018. Modelling the Trend of Zagros Forest Degradation using Logistic Regression (Case study: Chardavol Forest of Ilam province). Geography and Sustainability of Environment, 8 (27): 1- 13. (In Persian).
25. Mahdavi A. 2012. Forests and rangelands Wild fire risk zoning using GIS and AHP techniques. Caspian Journal of Environmental Sciences. 10 (1), 43- 52. (In Persian).
26. Meng Y, Deng Y, Shi P. 2015. Mapping forest wildfire risk of the world. World Atlas of Natural Disaster Risk, Springer, Berlin, Heidelberg: 261- 275. DOI:10.1007/978-3-662-45430-5_14
27. MohaJane M, Costache R, Karimi F, Baopham Q, Essahlaoui A, Nguyen H, Laneve G, Oudija F. 2021. Application of remote sensing and machine Learning algorithms for forest fire mapping in a Mediterranean area. Ecological Indicators, 129: 1- 17. https://doi.org/10.1016/j.ecolind.2021.107869
28. Mohammadi F, Shabanian N, Pourhashemi M, Fatehi P. 2011. Risk zone mapping of forest fire using GIS and AHP in a part of Paveh forests. Iranian Journal of Forest and Poplar Research, 18 (4): 569- 586. (In Persian).
29. Mohammadpoor M, Dashti S. 2020. Fire Simulation and Fire Spreading Zoning in Forest Ecosystem Using FARSITE Model (Case Study: Ilam Province Forests). Journal of Geography and Environmental Hazards, 8 (32): 87- 101. (In Persian).
30. Monavvari M. 2004. Guide to environmental impact assessment of tourism and nature tourism projects. Farzaneh Book Publications, 160 pp. (In Persian).
31. Oliveira S, Oehler F, San-Miguel-Ayanz, J, Camia, A, Pereira J.M.C. 2012. Modeling spatial patterns of fire occurrence in Mediterranean Europe using Multiple Regression and Random Forest. Forest Ecology and Management, 275: 117– 129. https://doi.org/10.1016/j.foreco.2012.03.003
32. Omidi M, Mafi Gholami D, Mahmoodi B, Jafari A. 2020. Spatial modeling the probability of wildfire occurrence using frequency ratio and weight- of-evidence models. Iranian Journal of Forest and Range Protection Research, 17 (2): 125- 144. (In Persian).
33. Organization of Agriculture- Jahad- Ilam. 2000. Physiographic report of DareShahr watershed, 150 pp. (In Persian).
34. Pham B.T, Jaafari A, Avand M, Al-Ansari N, Dinh Du T, Yen H.P.H, … Tuyen T.T, 2020. Performance evaluation of machine learning methods for forest fire modeling and prediction. Symmetry, 12 (6): 10- 12. Doi: https://doi.org/10.3390/sym12061022
35. Sadeghifar M, Beheshti Al Agha A, Pourreza M. 2017. Variability of Soil Nutrients and Aggregate Stability in Different Times after Fire in Zagros Forests (Case Study: Paveh Forests). Ecology of Iranian Forests, 4 (8): 19- 27. (In Persian).
36. Sakr G.E, Elhajj I.H. 2010. Artificial intelligence for forest fire prediction: A comparative study. Proceedings of the Sixth International Conference on Forest Fire Research, Coimbra, Portugal: 653- 661. Doi: https://scholarhub.balamand.edu.lb/handle/uob/983
37. Tinya F, Kovács B, Bidló A, Dima B, Király I, Kutszegi G, Lakatos F, Mag Z, Márialigeti S, Nascimbene J, Samu F. 2021. Environmental drivers of forest biodiversity in temperate mixed forests–A multi-taxon approach. Science of the Total Environment, 15;795:148720.
38. Vasilakos C, Kalabokidis K, Hatzopoulos J, Matsinos I. 2009. Identifying wildland fire ignition factors through sensitivity anaiysis of a neural network. J Natural hazards, 50 (1): 125- 143. Doi: https://doi.org/10.1007/s11069-008-9326-3
39. Veysi R, Fattahi B, Khosrobeigi S. 2022. Predicting and preparing a risk map of rangeland fires using random forest algorithms and support vector machine (Case study: Arak rangelands). Journal of Rangeland, 16 (2): 413- 426. (In Persian).
40. Viljur ML, Abella SR, Adámek M, Alencar JB, Barber NA, Beudert B, Burkle LA, Cagnolo L, Campos BR, Chao A, Chergui B. 2022. The effect of natural disturbances on forest biodiversity: an ecological synthesis. Biological Reviews, 97(5):1930-47.
41. Xiao X, Boles S, Frolking S, Salas W, Moore Iii B, Li C, He L, Zhao R. 2002. Observation of flooding and rice transplanting of paddy rice fields at the site to landscape scales in China using VEGETATION sensor data. International Journal of Remote Sensing, 23(15): 3009-3022.
42. Xie Y, Peng M. 2018. Forest fire forecasting using ensemble learning approaches. Neural Computing and Applications, 2018: 1- 10. https://doi.org/10.1007/s00521-018-3515-0
43. Yousefi A, Jalilvand H. 2010. Investigation of fire situation on forest and range areas of Mazandaran Province (Natural Resources Administration) from 1994 to 2007. The Second International Conference on Climate Change and Dendrochronology in Khazar Ecosystems, Sari. (In Persian).
44. Zhao Z, Li W, Ciais P, Santoro M, Cartus O, Peng S, Yin Y, Yue C, Yang H, Yu L, Zhu L. 2021. Fire enhances forest degradation within forest edge zones in Africa. Nature Geoscience, 14(7):479-83.
45. Zumbrunnen T, Pezzattic G.B, Menéndezd P, Bugmann H, Bürgia M, Conederac M. 2011. Weather and human impacts on forest fires: 100 years of fire history in two climatic regions of Switzerland. Forest Ecology and Management, 261: 2188- 2199.
