برآورد تغییرات سطح جنگل‌های هیرکانی جنوب دریای کاسپین با ترکیب داده‌های سنجش‌ازدور و روش‌های یادگیری ماشین

نوع مقاله : پژوهشی

نویسندگان

1 سازمان جنگل ها، مراتع و آبخیزداری، تهران، ایران

2 موسسه بین رشته ای اینترنت (IN3)، دانشگاه اوبرتا د کاتالونیا، بارسلون، کاتالونیا، اسپانیا

3 دانشجوی دکترای منابع آب، گروه علوم و مهندسی آب، دانشگاه بیرجند، بیرجند، ایران

4 دانشیار گروه علوم و مهندسی آب، دانشگاه بیرجند، بیرجند، ایران

5 دانشجوی دکترای، گرایش مهندسی آب و سازه های هیدرولیکی، دانشکده‌ مهندسی عمران، دانشگاه تهران، تهران، ایران

چکیده

جنگل‌های هیرکانی به دلیل قدمت چند میلیون ساله و تنوع زیستی غنی، در مناطق شمالی ایران و سواحل جنوبی دریای خزر از اهمیت زیست‌محیطی، اقتصادی و فرهنگی برخوردارند و نقش مهمی در حفظ منابع آب، خاک، تنوع گیاهی و جانوری، و تعدیل تغییرات اقلیمی ایفا می‌کنند. بررسی تغییرات پوشش جنگلی این جنگل‌ها برای مدیریت پایدار منابع طبیعی ضروری است. در این پژوهش، برای تحلیل دقیق تغییرات جنگل‌های هیرکانی بین سال‌های ۱۳۷۹ تا ۱۳۹۶، از داده‌های متنوع سنجش‌ازدور شامل شاخص تفاضل پوشش گیاهی نرمال‌شده (NDVI)، داده‌های پوشش گیاهی مادیس با عنوان (VCF)، و تصاویر ماهواره‌های سنتینل یک، لندست­های پنج و هشت استفاده شد. برای طبقه‌بندی، از روش‌های ماشین بردار پشتیبان (SVM) و جنگل تصادفی (RF) بهره گرفته شد. نتایج نشان داد که در بازه زمانی 17 ساله حدود 534 کیلومترمربع از مساحت جنگل‌های هیرکانی تخریب شده است. همچنین دقت کاربر برای  SVM، 93/26 و برای RF، 89/29 درصد بود. ضریب کاپا نیز برای SVM ،  94/62 و برای  RF، 74/63 درصد به دست آمد که حاکی از دقت بالای نتایج به‌دست‌آمده است. مقایسه نتایج میزان تغییرات پوشش جنگل و دقت به‌دست‌آمده حاصل از طبقه‌بندی با الگوریتم‌ها نشان داد که نتایج با تقریب خوبی با داده‌های جهانی تغییر پوشش جنگل هانسن مطابق دارد. رویکرد بکار رفته در این تحقیق و نتایج آن می‌تواند در مدیریت و برنامه‌ریزی حفاظت از جنگل‌ها و مدیریت منابع طبیعی به‌کار گرفته شود و به بهره‌برداری پایدار از جنگل‌های هیرکانی کمک شایانی کند.

کلیدواژه‌ها

عنوان مقاله [English]

Using Remote Sensing Data and Machine Learning Methods to Estimate Changes in Hyrcanian Forests along the Southern Coasts of the Caspian Sea

نویسندگان [English]

  • Mahdi Afraz 1
  • Davoud Omarzadeh 2
  • Mobin Eftekhari 3
  • Mostafa Yaghoobzadeh 4
  • Ali Haji Elyasi 5
1 Forests, Range and Watershed Management Organization, Tehran, Iran.
2 Internet Interdisciplinary Institute (IN3), Universitat Oberta de Catalunya, Barcelona, Catalonia, Spain
3 Ph.D Student, Water Engineering Department, University of Birjand, Birjand, Iran
4 Associate Professor, Department of Water Engineering, University of Birjand, Birjand, Iran.
5 Ph.D. Candidate, School of Civil Engineering, College of Engineering, University of Tehran, Tehran, Iran
چکیده [English]

The Hyrcanian Forests, located along the southern coastal areas of the Caspian Sea in northern Iran, are of great environmental, economic, and cultural significance. They play crucial roles not only in preserving water resources, soil, plant, and animal diversity but in mitigating adverse impacts of climate change as well. The present study investigated changes in the Hyrcanian forest cover between 2000 and 2017 using the diverse remote sensing data of Normalized Difference Vegetation Index (NDVI), and MODIS Vegetation Continuous Fields (VCF) as well as Sentinel-1, Landsat-5, and Landsat-8 satellite images while the Support Vector Machine (SVM) and Random Forest (RF) methods were employed for classification. The results revealed that approximately 534 square kilometers of the forests had experienced degradation. Moreover, classification accuracy levels were impressive as evidenced by a user accuracy of 93.26% and a Kappa coefficient of 94.62% recorded for SVM and corresponding values of 89.29% 74.63% for RF. Comparison with global forest change datasets confirmed the reliability of the results obtained. The research approach seems to offer promising insights useful for forest conservation management, natural resource planning, and enhanced sustainable utilization of Hyrcanian forests.

کلیدواژه‌ها [English]

  • Forest degradation
  • remote sensing
  • Hansen method
  • Support Vector Machine
  • Random Forest

مراجع

  1. Ahmad, A., Ahmad, S., Nabi, G., Liu, Q.J., Islam, N. and Luan, X., 2022. Trends in deforestation as a response to management regimes and policy intervention in the Hindu Kush Himalaya of Pakistan. Frontiers in Environmental Science, 10, p.810806. DOI: 10.3389/fenvs.2022.810806
  2. Akike, S. and Samanta, S., 2016. Land use/land cover and forest canopy density monitoring of Wafi-Golpu project area, Papua New Guinea. Journal of Geoscience and Environment Protection, 4, pp.1-14. http://dx.doi.org/10.4236/gep.2016.48001
  3. Aziz, G., Minallah, N., Saeed, A., Frnda, J. and Khan, W., 2024. Remote sensing-based forest cover classification using machine learning. Scientific Reports, 14(1), p.69.

DOI: 10.1038/s41598-023-50863-1

  1. Beygi Heidarlou, H., Banj Shafiei, A., Nasiri, V., Niţă, M.D., Borz, S.A. and Lopez-Carr, D., 2023. Impact of Iran’s forest nationalization law on forest cover changes over six decades: A case study of a zagros sparse coppice oak forest. Sensors, 23(2), p.871.

DOI: 10.3390/s23020871

  1. Silva Junior, C.H.L., Pessôa, A.C.M., Carvalho, N.S., Reis, J.B.C., Anderson, L.O. and Aragão, L.E.O.C., 2021. The Amazon’s deforestation rate in 2020 is projected to be the greatest of the decade. Nature Ecology & Evolution, 5, pp.144-145. DOI:1038/s41559-020-01368-x
  2. Chowdhury, M.S., 2023. Comparison of accuracy and reliability of random forest, support vector machine, artificial neural network and maximum likelihood method in land use/cover classification of urban setting. Environmental Challenges, p.100800.

DOI:10.1016/j.envc.2023.100800

  1. Congalton, R.G., 1991. A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment, 37(1), pp.35-46. DOI:1016/0034-4257(91)90048-B
  2. Coops, N.C., Tompalski, P., Goodbody, T.R., Achim, A. and Mulverhill, C., 2023. Framework for near real-time forest inventory using multi source remote sensing data. Forestry, 96(1), pp.1-19. DOI:1093/forestry/cpac015
  3. Darvishsefat, A. and Shirvani, A., 2019. Monitoring and modeling changes of forest area using logistic regression-markov and geomod. Geographic Space, 19(65), pp.171-189. (In Persian).
  4. Ebrahimy, H., Naboureh, A., Feizizadeh, B., Aryal, J. and Ghorbanzadeh, O., 2021. Integration of Sentinel-1 and Sentinel-2 data with the G-SMOTE technique for boosting land cover classification accuracy. Applied Sciences, 11(21), p.10309. DOI:10.3390/app112110309
  5. FAO, 2020. Global Forest Resources Assessment 2020. Report Iran. Rome: FAO
  6. Feizizadeh, B., Lakes, T., Omarzadeh, D., Sharifi, A., Blaschke, T. and Karimzadeh, S., 2022. Scenario-based analysis of the impacts of lake drying on food production in the Lake Urmia Basin of Northern Iran. Scientific Reports, 12(1), p.6237. DOI: 10.1038/s41598-022-10228-6
  7. Feizizadeh, B., Omarzadeh, D., Mohammadnejad, V., Khallaghi, H., Sharifi, A. and Karkarg, B.G., 2023. An integrated approach of artificial intelligence and geoinformation techniques applied to forest fire risk modeling in Gachsaran, Iran. Journal of Environmental Planning and Management, 66(6), pp.1369-1391. DOI: 10.1080/09640568.2022.2029347
  8. Forkuor, G., Dimobe, K., Serme, I. and Tondoh, J.E., 2017. Landsat-8 vs. Sentinel-2: examining the added value of Sentinel-2’s red-edge bands to land-use and land-cover mapping in Burkina Faso. GIScience & Remote Sensing, 55(3), pp.331-354. DOI: 10.1080/15481603.2017.1370169
  9. Gu, F., Ramezani, E., Alizadeh, K. and Behling, H., 2021. Vegetation dynamics, environmental changes, and anthropogenic impacts on the coastal Hyrcanian Forests in Northern Iran. Journal of Coastal Research, 37(3), pp.611-619. DOI: 10.2112/JCOASTRES-D-20-00067.1
  10. Haidari, M. and Karamdoost Maryan, B., 2016. The study and policy cessation of forest utilization or forest logging in Hyrcanian forests (Case study: Shafarood forests). Iranian Journal of Forest and Poplar Research, 24(4), pp.724-736.

 https://doi.org/10.22092/ijfpr.2016.109450

  1. Hansen, M.C., Potapov, P.V., Moore, R., Hancher, M., Turubanova, S.A., Tyukavina, A., Thau, D., Stehman, S.V., Goetz, S.J., Loveland, T.R., Kommareddy, A., Egorov, A., Chini, L., Justice, C.O. and Townshend, J.R.G., 2013. High-resolution global maps of 21st-century forest cover change. Science, 342(6160), pp.850-853. DOI: 10.1126/science.1244693
  2. Hashemi, S.A., Fatemi Talab, S.R., Kavousi Kalashmi, H. and Madanipour Kermanshahi, M., 2016. Change detection in the forest cover of Siyahmezgi watershed of Guilan using LandSat images. Journal of RS and GIS for Natural Resources, 7(3), pp.78-88. (In Persian).
  3. Hassan, M.M., Duveneck, M. and Southworth, J., 2023. The role of the refugee crises in driving forest cover change and fragmentation in Teknaf, Bangladesh. Ecological Informatics, 74, p.101966. DOI: 10.1016/j.ecoinf.2022.101966
  4. Hu, L., Shariff, A.R.B.M., Omar, H., Song, D.X. and Wu, H., 2022. GEE-based spatiotemporal evolution of deforestation monitoring in Malaysia and its drivers. In: Remote Sensing Application: Regional Perspectives in Agriculture and Forestry. Singapore: Springer Nature Singapore, pp.279-302. DOI: 10.1007/978-981-19-7620-3_13
  5. Huang, C. and Zhang, C., 2022. Characterizing urban growth in Vientiane from 2000 to 2019 using time-series optical and SAR-based estimates of urban land. International Journal of Applied Earth Observation and Geoinformation, 109, p.102798.

DOI: 10.1016/j.jag.2022.102798

  1. Jourgholami, M. and Majnounian, B., 2011. Harvesting systems in Hyrcanian Forest, Iran; limitations and approaches.
  2. Karimzadeh Jafari, E., Naghavi, H., Adeli, K. and Latifi, H., 2020. A nondestructive, remote sensing-based estimation of the economic value of aboveground temperate Forest biomass (case study: Hyrcanian Forests, Nowshahr-Iran). Journal of Sustainable Forestry, 39(7), pp.750-770. DOI: 10.1080/10549811.2020.1723113
  3. Li, H., Wan, W., Fang, Y., Zhu, S., Chen, X., Liu, B. and Hong, Y., 2019. A Google Earth Engine-enabled software for efficiently generating high-quality user-ready Landsat mosaic images. Environmental Modelling & Software, 112, pp.16-22.

DOI: 10.1016/j.envsoft.2018.11.004

  1. Liu, P., Choo, K.K.R., Wang, L. and Huang, F., 2017. SVM or deep learning? A comparative study on remote sensing image classification. Soft Computing, 21, pp.7053-7065.

DOI: 10.1007/s00500-016-2247-2

  1. Lynch, J., Maslin, M., Balzter, H. and Sweeting, M., 2013. Choose satellites to monitor deforestation. Nature, 496(7445), pp.293-294. DOI: 10.1038/496293a
  2. Manfreda, S., McCabe, M.F., Miller, P.E., Lucas, R., Pajuelo Madrigal, V., Mallinis, G. and Toth, B., 2018. On the use of unmanned aerial systems for environmental monitoring. Remote Sensing, 10(4), p.641. DOI: 10.3390/rs10040641
  3. Mehmood, M., Shahzad, A., Zafar, B., Shabbir, A. and Ali, N., 2022. Remote sensing image classification: A comprehensive review and applications. Mathematical Problems in Engineering, 2022(1), p.5880959. DOI: 10.1155/2022/5880959
  4. Mirakhorlu and Akhwan, 2017. Evaluation of changes in the level of Hyrcanian forests from 1383 to 1395. Nature of Iran, 2(3), pp.40-45. (In Persian).
  5. Mohammadpour, P., Viegas, D.X. and Viegas, C., 2022. Vegetation mapping with random forest using Sentinel 2 and GLCM texture feature—A case study for Lousã region, Portugal. Remote Sensing, 14(18), p.4585. DOI: 10.3390/rs14184585
  6. Moharrami, M., Attarchi, S., Gloaguen, R. and Alavipanah, S.K., 2024. Monitoring land cover changes in northwestern Iran using training samples migration method. Iranian Journal of Remote Sensing & GIS, 16(1), pp.129-154. doi:10.48308/gisj.2024.233836.1188
  7. Nakhutsrishvili, G., Zazanashvili, N., Batsatsashvili, K. and Montalvo, C.S., 2015. Colchic and Hyrcanian forests of the Caucasus: Similarities, differences and conservation status. Flora Mediterranea, 25(Special Issue), pp.185-192. DOI: 10.7320/FlMedit25SI.185
  8. Naqinezhad, A. and Zarezadeh, S., 2013. Phytosociological survey of Noor and Sisangan lowland Hyrcanian forests Mazandaran province. Journal of Plant Biological Sciences, 5(16), pp.103-121. (In Persian).
  9. Nasiri, V., Beloiu, M., Darvishsefat, A.A., Griess, V.C., Maftei, C. and Waser, L.T., 2023. Mapping tree species composition in a Caspian temperate mixed forest based on spectral-temporal metrics and machine learning. International Journal of Applied Earth Observation and Geoinformation, 116, p.103154. DOI: 10.1016/j.jag.2022.103154
  10. Nasiri, V., Heidarlou, H.B., Alchin, A.A., Moradi, F., Rahmanian, S., Afshari, S. and Griess, V.C., 2023. How do conservation policies, climate and socioeconomic changes impact Hyrcanian forests of northern Iran? Ecological Informatics, 78, p.102351.

DOI: 10.1016/j.ecoinf.2023.102351

  1. Omarzadeh, D., Afraz, M., Akbari, M., Eftekhari, M. and Noghani, Z., 2021. Evaluation of changes in the Forest Environment in Guilan Province using a combination of Remote Sensing Data. The Malaysian Forester, 84(1), pp.65-83.
  2. Omarzadeh, D., González-Godoy, A., Bustos, C., Martín-Fernández, K., Scotto, C., Sánchez, C. and Borge-Holthoefer, J., 2024. Explainable automatic detection of fiber–cement roofs in aerial RGB images. Remote Sensing, 16(8), p.1342. DOI: 10.3390/rs16081342
  3. Scullion, J.J., Vogt, K.A., Drahota, B., Winkler-Schor, S. and Lyons, M., 2019. Conserving the last great forests: A meta-analysis review of the drivers of intact forest loss and the strategies and policies to save them. Frontiers in Forests and Global Change, 2, p.62.

DOI: 10.3389/ffgc.2019.00062

  1. Sefidi, K. and Esmaeilpour, M., 2022. Meta-analysis of snag volume variable in Hyrcanian forests. Journal of Wood and Forest Science and Technology, 29(2), pp.121-136. DOI:10.22069/jwfst.2022.20263.1971. (In Persian).
  2. Sharif, M. and Attarchi, S., 2024. Estimation of Hyrcanian forests coverage using ALOS PALSAR and MODIS satellite data (2007–2017). Advances in Space Research, 73(1), pp.426-439. DOI: 10.1016/j.asr.2023.09.033
  3. Sheykhmousa, M., Mahdianpari, M., Ghanbari, H., Mohammadimanesh, F., Ghamisi, P. and Homayouni, S., 2020. Support vector machine versus random forest for remote sensing image classification: A meta-analysis and systematic review. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13, pp.6308-6325.

DOI: 10.1109/JSTARS.2020.3026724

  1. Sotirov, M., Pokorny, B., Kleinschmit, D. and Kanowski, P., 2020. International forest governance and policy: institutional architecture and pathways of influence in global sustainability. Sustainability, 12(17), p.7010. DOI: 10.3390/su12177010
  2. Stritih, A., Bebi, P., Rossi, C. and Grêt-Regamey, A., 2021. Addressing disturbance risk to mountain forest ecosystem services. Journal of Environmental Management, 296, p.113188. DOI: 10.1016/j.jenvman.2021.113188
  3. Tariq, A., Yan, J., Gagnon, A.S., Riaz Khan, M. and Mumtaz, F., 2023. Mapping of cropland, cropping patterns and crop types by combining optical remote sensing images with decision tree classifier and random forest. Geo-Spatial Information Science, 26(3), pp.302-320.

DOI: 10.1080/10095020.2022.2100287

  1. Tohidifar, M., Moser, M., Zehzad, B. and Ghadirian, T., 2016. Biodiversity of the Hyrcanian forests: A synthesis report.
مدیریت اراضی
دوره 12، شماره 2
اسفند 1403
صفحه 169-185

فایل ها

هم رسانی

ارجاع به این مقاله

آمار