پیش‌بینی حساسیت زمین‌لغزش با استفاده از مدل‌های ترکیبی فاصله ماهالانوبیس و یادگیری ماشین (مطالعه موردی: حوزه آبخیز اوغان، استان گلستان)

نوع مقاله : علمی -پژوهشی

نویسندگان

1 گروه آبخیزداری، دانشکده مرتع و آبخیزداری، دانشگاه علوم کشاورزی و منابع طبیعی گرگان، گرگان، ایران

2 گروه مهندسی منابع طبیعی و محیط زیست، دانشکده کشاورزی، دانشگاه شیراز، شیراز، ایران

3 دانشگاه SNHU، نیوهمپشیر، آمریکا

چکیده

هدف از تحقیق پیش‌رو، پهنه‌بندی حساسیت زمین‏لغزش در حوزه‌ آبخیز اوغان، واقع در استان گلستان می­باشد. بدین منظور از دو مدل توانمند داده­کاوی شامل جنگل تصادفی و بیشینه آنتروپی استفاده گردید. زمین­لغزش­ها با استفاده از الگوریتم فاصله ماهالانوبیس به دو دسته 70 درصد (واسنجی پارامترها و تعلیم مدل­ها) و 30 درصد (اعتبارسنجی نتایج مدل­ها) تقسیم شدند. هم­چنین با توجه به‌ مرور منابع گسترده، 15 عامل مؤثر بر وقوع زمین­لغزش در منطقه موردمطالعه با روش تورم واریانس غربال، عوامل بهینه انتخاب و لایه­های رقومی عوامل در سامانه اطلاعات جغرافیایی تهیه شدند. به­منظور ارزیابی نتایج مدل­ها (قدرت یادگیری و اعتبارسنجی نتایج) از مقدار مساحت زیرمنحنی تشخیص عملکرد نسبی با استفاده از دو دسته داده واسنجی و اعتبارسنجی استفاده شد. نتایج حاصل از ارزیابی قدرت یادگیری مدل­ها نشان داد که مدل جنگل تصادفی و بیشینه آنتروپی به ­ترتیب با مقادیر سطح زیر منحنی 923/0 و 91/0 دارای قدرت یاگیری و برازش نسبتاً مشابهی می­باشند. اگرچه در مرحله اعتبارسنجی مشخص گردید که مدل جنگل تصادفی با مقدار 9/0 نسبت به مدل بیشینه آنتروپی با مقدار 85/0 قدرت پیش­بینی و تعمیم نتایج بالاتری دارد. لذا مدل جنگل تصادفی به‌ عنوان مدل برتر در ارزیابی حساسیت زمین­لغزش حوزه آبخیز اوغان معرفی گردید. براساس نتایج مدل جنگل تصادفی، حدود 10 درصد از حوزه آبخیز اوغان در پهنه حساسیت زیاد و خیلی­زیاد به­وقوع زمین­لغزش قرار گرفته است. هم­چنین، عوامل بارش، شاخص تفاضلی پوشش گیاهی نرمال شده، شاخص ارتفاع از سطح نزدیک‌ترین زهکش، سنگ­شناسی و فاصله از جاده به‌عنوان مهم­ترین عوامل مؤثر در وقوع زمین­لغزش­های منطقه معرفی گردیدند.

کلیدواژه‌ها


عنوان مقاله [English]

(Landslide susceptibility prediction using the coupled Mahalanobis distance and machine learning models (case study: Owghan watershed, Golestan province

نویسندگان [English]

  • Aiding Kornejady 1
  • Majid Ownegh 1
  • Hamid Reza Pourghasemi 2
  • Abdolreza Bahremand 1
  • Manouchehr Motamedi 3
1 Department of Watershed Management, Faculty of Rangeland and Watershed Management, Gorgan University of Agricultural Science and Natural Resources (GUASNR), Gorgan, Iran.
2 Department of Natural Resources and Environmental Engineering, College of Agriculture, Shiraz University, Shiraz, Iran.
3 Southern New Hampshire University, New Hampshire
چکیده [English]

IntroductionLandslide susceptibility maps are considered a backbone for decision-makers to suggest solitary or combined technical and regulatory measures. Such maps are also considered an invaluable tool for engineers, earth scientists, planners, and decision-makers to select the most suitable areas for agriculture, building, and other development activities. Hence, thanks to landslide susceptibility maps, addressing highly susceptible areas are feasible, so that over the course of further detailed studies on the imminent landslide occurrences in the future, landslide potential risk is mitigated.Materials and methodsIn this study, two robust data mining models, namely random forest and maximum entropy were used to map landslide susceptibility across the Owghan Watershed in Golestan province. After preparing the landslide inventory map via extensive field surveys, interpreting Google Earth images, and the archived data acquired from different organizations, landslide points were split into two sets of training (70%) and validation (30%) by using the Mahalanobis distance technique. Further, drawing on the extensive literature review, fifteen factors including climatic, geological, tectonic, topo-hydrological, and anthropogenic drivers, as landslide-controlling factors were selected and sieved through the variance inflation factor test. Ultimately, by implementing the above-mentioned data mining techniques, the most important factors in the modeling process, as well as the highly susceptible locations in the study area, were introduced.Results and discussionEvaluating the learning capability, both the random forest and maximum entropy models with the respective area under the receiver operating characteristic curve (AUROC) values of 0.923 and 0.91, showed almost identical fitting abilities. However, getting to the validation stage, it was found that the random forest with the AUROC value of 0.9 clearly outperforms maximum entropy (AUROC= 0.85) in terms of prediction power and generalization capacity. Hence, the random forest was suggested as a better-performing model for landslide susceptibility mapping in the Owghan watershed, compared to its counterpart. About 10% of the study area falls into high and very high landslide susceptibility zones. Furthermore, five landslide-controlling factors including rainfall, normalized difference vegetation index, height above the nearest drainage, lithological formation, and proximity to roads have been found to be the most significant factors contributing to landslide occurrence in the study area. Additionally, the results attest that announcing the Safiabad village as a landslide-prone area by the authorities is technically sound and evacuating the residents to a new place has been a right decision; however, some parts of the newly inhabited area shows landslide predisposing patterns which can lead to a higher susceptibility of the area to landslide occurrence in the future.ConclusionScrutinizing the results of random forest model revealed that a combination of natural factors (intense rainfall, bare lands, susceptible lithological formations, and topo-hydrological mechanisms) and anthropogenic interferences (tillage parallel to slope length/perpendicular to contour lines and unprincipled road construction) are synergistically responsible for landslide occurrence in the Owghan Watershed. On the other hand, announcing the Safiabad village as a critical landslide-prone area seems to be a wise decision, although the newly inhabited place seems to be selected merely based on having a suitable slope steepness (i.e., almost flat) and being accessible through several connecting routes, while the enhanced conservation tillage methods have not been applied to the selected site and adjacent areas. The latter, according to our inferences, can trigger a crisis in a larger extent. Moreover, owing to the presence of other landslide predisposing factors in the new residential site, safe areas should be pointed out and announced by adopting a holistic view on the entire influential and predisposing conditions for landslide occurrence.

کلیدواژه‌ها [English]

  • Maximum Entropy
  • Random forest
  • Geographical information system
  • Machine learning models
  1. -اداره کل منابع طبیعی استان گلستان (مدیریت آبخیزداری)، 1388. مطالعات سازه‌های کنترل سیل و رسوب حوزه آبخیز اوغان-گالیکش، شرکت مهندسی مشاور کاوش پی مشهد، مطالعات پایه.
  2. -بهره‌مند، ع. و کرنژادی، آ.، 1394. معرفی و تهیه شاخص جدید توپوگرافی ارتفاع از سطح نزدیکترین زهکش (HAND). کنفرانس و نمایشگاه مهندسی آب با محوریت تجاری‌سازی، مرکز همایش‌های بین‌المللی شهید بهشتی.
  3. -بهره‌مند، ع.ب. و کرنژادی، آ.، 1394. معرفی شاخص جدید توپوگرافیکی-هیدرولوژیکی ارتفاع از سطح نزدیکترین زهکش (HAND) و ترسیم نقشه HAND برای حوزه چهل‌چای، همایش ملی پژوهش‌های نوین در مدیریت منابع طبیعی، دانشگاه ملایر، دانشکده منابع طبیعی و محیط‌زیست، گروه مرتع و آبخیزداری.
  4. -طالبی، ع.، گودرزی، س. و پورقاسمی، ح.، 1397. بررسی امکان تهیه نقشه خطر زمین‌لغزش با استفاده از الگوریتم جنگل تصادفی (محدوده مورد مطالعه: حوزه آبخیز سردارآباد، استان لرستان) دوره 7، شماره 16 - شماره پیاپی 2، ص 45-64.
  5. -محمدی، م. و پورقاسمی، ح.، 1396. اولویت‌بندی عوامل موثر بر وقوع زمین‌لغزش و تهیه نقشه حساسیت آن با استفاده از الگوریتم نوین جنگل تصادفی (مطالعه موردی: بخشی از استان گلستان)، پژوهشنامه مدیریت حوزه آبخیز، سال 8، شماره 15، ص 161-170.
  6.  
  7.  
  8.  
  9. -Akgun, A., 2012. A comparison of landslide susceptibility maps produced by logistic regression, multi-criteria decision, and likelihood ratio methods: a case study at İzmir, Turkey: Landslides, v. 9, p. 93-106.
  10. -Ambraseys, N.N. and Melville, C.P., 1982. A History of Persian Earthquakes: Cambridge Univ, Press, New York, 219 p.
  11. -Ayalew, L. and Yamagishi, H., 2005. The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan: Geomorphology, v. 65, p. 15-31.
  12. -Beven, K.J. and Kirkby, M.J., 1979. A physically based, variable contributing area model of basin hydrology/Un modèle à base physique de zone d'appel variable de l'hydrologie du bassin versant: Hydrological Sciences Journal, v. 24, p. 43-69.
  13. -Chen, W., Pourghasemi, H.R., Kornejady, A. and Zhang, N., 2017. Landslide spatial modeling: introducing new ensembles of ANN, MaxEnt, and SVM machine learning techniques: Geoderma, v. 305, p. 314-327.
  14. -Elith, J., Graham, C.H., Anderson, R.P., Dudik, M., Ferrier, S., Guisan, A. and Li, J., 2006. Novel methods improve prediction of species’ distributions from occurrence data: Ecography, v. 29, p. 129-151.
  15. -Ercanoglu, M. and Gokceoglu, C., 2002. Assessment of landslide susceptibility for a landslide-prone area (north of Yenice, NW Turkey) by fuzzy approach: Environmental Geology, v. 41, p. 720-730.
  16. -Fukunaga, K., 2013. Introduction to statistical pattern recognition: Academic Press, Boston, 592 p.
  17. -Jenness, J., Brost, B. and Beier, P., 2013. Land facet corridor designer: USDA forest service rocky mountain research station.
  18. -Joyner, W.B. and Boore, D.M., 1981. Peak horizontal acceleration and velocity from strong-motion records including records from the 1979 Imperial Valley, California, earthquake: Bulletin of the Seismological Society of America, v. 71, p. 2011-2038.
  19. -Kornejady, A., Ownegh, M. and Bahremand, A., 2017a. Landslide susceptibility assessment using maximum entropy model with two different data sampling methods: Catena, v. 152, p. 144-162.
  20. -Kornejady, A., Ownegh, M., Rahmati, O. and Bahremand, A., 2018b. Landslide susceptibility assessment using three bivariate models considering the new topo-hydrological factor: HAND: Geocarto International, v. 33, p. 1155-1185.
  21. -Kornejady, A., Pourghasemi, H.R. and Afzali, S.F., 2019. Presentation of RFFR New Ensemble Model for Landslide Susceptibility Assessment in Iran, In Landslides: Theory, Practice and Modelling: Springer, Cham, p. 123-143.
  22. -Lee, S., Hwang, J. and Park, I., 2013. Application of data-driven evidential belief functions to landslide susceptibility mapping in Jinbu, Korea: Catena, v. 100, p. 15-30.
  23. -Meten, M., PrakashBhandary, N. and Yatabe, R., 2015. Effect of Landslide Factor Combinations on the Prediction Accuracy of Landslide Susceptibility Maps in the Blue Nile Gorge of Central Ethiopia: Geoenvironmental Disasters, v. 2, p. 1-17.
  24. -O’brien, R.M., 2007. A caution regarding rules of thumb for variance inflation factors: Quality & Quantity, v. 41, p. 673-690.
  25. -Phillips, S.J., Anderson, R.P. and Schapire, R.E., 2006. Maximum entropy modeling of species geographic distributions: Ecological Modelling, v. 190, p. 231-259.
  26. -Pontius Jr, R.G. and Schneider, L.C., 2001. Land-cover change model validation by an ROC method for the Ipswich watershed, Massachusetts, USA: Agriculture, Ecosystems & Environment, v. 85, p. 239-248.
  27. -Pourghasemi, H.R. and Rahmati, O., 2018. Prediction of the landslide susceptibility: Which algorithm, which precision?: Catena, v. 162, p. 177-192.
  28. -Radbruch-Hall, D.H. and Varnes, D.J., 1976. Landslides—cause and effect: Bulletin of the International Association of Engineering Geology-Bulletin de l'Association Internationale de Geologie de l'Ingenieur, v. 13, p. 205-216.
  29. -Rahmati, O., Kornejady, A., Samadi, M., Nobre, A.D. and Melesse, A.M., 2018. Development of an automated GIS tool for reproducing the HAND terrain model: Environmental Modelling & Software, v. 102, p. 1-12.
  30. -Shataee, S., Kalbi, S., Fallah, A. and Pelz, D., 2012. Forest attribute imputation using machine-learning methods and ASTER data: comparison of k-NN, SVR and random forest regression algorithms: International Journal of Remote Sensing, v. 33, p. 6254-6280.
  31. -Stumpf, A. and Kerle, N., 2011. Object-oriented mapping of landslides using Random Forests: Remote Sensing of Environment, v. 115, p. 2564-2577.
  32. -Tien Bui, D.T., Lofman, O., Revhaug, I. and Dick, O., 2011. Landslide susceptibility analysis in the Hoa Binh province of Vietnam using statistical index and logistic regression: Natural Hazards, v. 59, p. 1413.
  33. -Trigila, A., Frattini, P., Casagli, N., Catani, F., Crosta, G., Esposito, C. and Spizzichino, D., 2013. Landslide susceptibility mapping at national scale: the Italian case study: In Landslide science and practice, p. 287-295: Springer, Berlin, Heidelberg.
  34. -Tsangaratos, P. and Benardos, A., 2014. Estimating landslide susceptibility through a artificial neural network classifier: Natural Hazards, v. 74, p. 1489-1516.
  35. -United Nations Office for Disaster Risk Reduction, 2005. National report of the Islamic Republic of Iran: World Conference on Disaster Reduction, Kobe, Hyogo, Japan.
  36. -Van der Knijff, J.M., Jones, R.J.A. and Montanarella, L., 2000. Soil erosion risk assessment in Europe.
  37. -Van Westen, C.J., Van Asch, T.W. and Soeters, R., 2006. Landslide hazard and risk zonation—why is it still so difficult?: Bulletin of Engineering Geology and the Environment, v. 65, p. 167-184.
  38. -Vorpahl, P., Elsenbeer, H., Märker, M. and Schröder, B., 2012. How can statistical models help to determine driving factors of landslides?: Ecological Modelling, v. 239, p. 27-39.
  39. -Wald, D.J., Quitoriano, V., Heaton, T.H. and Kanamori, H., 1999. Relationships between peak ground acceleration, peak ground velocity, and modified Mercalli intensity in California: Earthquake Spectra, v. 15, p. 557-564.