İstanbul İçin CBS Tabanlı Makine Öğrenmesi İle Sel Duyarlılık Haritasının Oluşturulması

Öz

Bu çalışma kapsamında meydana gelebilecek olası bir sel olayının gerçekleşebileceği yerin önceden tahmini ve tespiti için makine öğrenmesi yöntemleri kullanılarak coğrafi bilgi sistemleri (CBS) tabanlı bir sel duyarlılık haritalama modeli oluşturulması amaçlanmıştır. Çalışma kapsamında incelen bölge olarak ise Türkiye’nin metropol kenti olan İstanbul ili seçilmiştir. Literatürden elde edilen sel envanteriyle oluşturulan örneklem kümesi önce sel olmayan noktaların rastgele oluşturulması ile genişletilmiş olup, ardından sınıf dengesizliği rastgele alt örnekleme (RUS) tekniği ile giderilmiştir. Bu yaklaşım Türkiye’ de gerçekleştirilen sel duyarlılık haritalamaları çalışmaları için ilk kez uygulanmıştır. Rastgele orman (RF), stokastik gradyan artırma (SGB) ve XGBoost algoritmaları olmak üzere üç farklı makine öğrenmesi algoritmasının performans karşılaştırmaları gerçekleştirilmiştir. En yüksek model performansının XGBoost ile elde edildiği, bu metodu ise sırasıyla SGB ve RF’nin takip ettiği sonucuna ulaşılmıştır. Ayrıca, RF ve SGB modellerinin sel olmayan noktaların neredeyse tamamını doğru olarak bulduğu, sel olan noktalarda ise %90.67’lik bir başarı sergilediği görülmüştür. Fakat, çalışmanın esas amacını kapsayan sel gerçekleşen noktaların belirlenmesinde XGBoost modeli %92.00’lik bir başarı ile diğer iki metoda üstünlük sergilediği tespit edilmiştir. Sel olayını etkileyen parametreler incelendiğinde ise İstanbul için seli en önemli parametrenin yağış olduğu sonucuna ulaşılmış olup, yağışı sırasıyla drenaj ağına uzaklık ve eğri numarası takip etmiştir. Sonuç olarak çalışma kapsamında İstanbul’da gerçekleştirilen sel duyarlılık haritalamaları çalışmaları için ilk kez uygulanan bu çerçevenin kullanımının sayısı ve etkileri giderek artırılarak sel olaylarına karşı daha yaygın alanlara uygulanması gelecek vadedici bir yaklaşım olacaktır.

Anahtar Kelimeler

• Coğrafi Bilgi Sistemleri
• İstanbul
• Makine Öğrenmesi
• Risk Yönetimi
• Sel Duyarlılık Haritalaması
• Uzaktan Algılama

Generating the Flood Susceptibility Map for Istanbul with GIS-Based Machine Learning Algorithms

Öz

The main objective of the current study is to generate a flood hazard map by using the machine learning algorithms hybridized with the geographic information systems (GIS). In this regard, the province of Istanbul, which is the metropolitan city of Turkey, was selected as the focal region within the scope of the study. The class imbalance was tackled through the commonly used random under sampling (RUS) technique in order to create a fair comparison datum line. It is worth mentioning that this is the first time this approach has been used for flood hazard mapping studies in Turkey. Random forest (RF), stochastic gradient boosting (SGB), and XGBoost algorithms were used. The best predictive performance was obtained with the XGBoost algorithm, followed by SGB and RF, respectively. The RF and SGB models showed a 90.67% success rate in determining the inundation points, while the XGBoost model outperformed its counterparts with a 92.00% success rate in determining the inundation points. In this research, the importance levels of the flood triggering variables were further investigated in order to enliven the comprehensibility of the obtained results. Thus, the most important variable was the precipitation, followed by the distance to the drainage network and the number of curves, respectively. Finally, it is suggested that flood vulnerability mapping attempts can be considered as promising approaches against increasing flood incidents over the years.

Anahtar Kelimeler

• Geographical Information Systems
• Istanbul
• Machine Learning
• Risk Management
• Flood Hazard Mapping
• Remote Sensing

Kaynakça

1. Abedi, R., Costache, R., Shafizadeh-Moghadam, H., & Pham, Q. B. (2021). Flash-flood susceptibility mapping based on XGBoost, random forest and boosted regression trees. Geocarto International, 37(19), 5479–5496. https://doi.org/10.1080/10106049.2021.1920636
2. Anılan, T., Durmuş, H., Akçalı, E., & Yüksek, M. (2021). Taşkın farkındalık ve erken uyarı sistemleri değerlendirmesi: Trabzon Beşikdüzü örneği. Doğal Afetler ve Çevre Dergisi, 7(1), 110–123. https://doi.org/10.21324/dacd.722798
3. Avand, M., Khiavi, A. N., Khazaei, M., & Tiefenbacher, J. P. (2021). Determination of flood probability and prioritization of sub-watersheds: A comparison of game theory to machine learning. Journal of Environmental Management, 295, Article 113040. https://doi.org/10.1016/j.jenvman.2021.113040
4. Aydin, H. E., & Iban, M. C. (2022). Predicting and analyzing flood susceptibility using boosting-based ensemble machine learning algorithms with SHapley Additive exPlanations. Natural Hazards, 116(3), 2957–2991. https://doi.org/10.1007/s11069-022-05793-y
5. Başakın, E. E., Ekmekcioğlu, M., Çıtakoğlu, H., & Özger, M. (2021). A new insight to the wind speed forecasting: robust multi-stage ensemble soft computing approach based on pre-processing uncertainty assessment. Neural Computing and Applications, 34(1), 783–812. https://doi.org/10.1007/s00521-021-06424-6
6. Bentéjac, C., Csörgő, A., & Martínez-Muñoz, G. (2020). A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54(3), 1937–1967. https://doi.org/10.1007/s10462-020-09896-5
7. Bhattacharya, S., S, S. R. K., Maddikunta, P. K. R., Kaluri, R., Singh, S., Gadekallu, T. R., Alazab, M., & Tariq, U. (2020). A Novel PCA-Firefly Based XGBoost Classification Model for Intrusion Detection in Networks Using GPU. Electronics, 9(2), Article 219. https://doi.org/10.3390/electronics9020219
8. Breiman, L. (2001). Random forests. Machine Learning, 45, 5–32.

9. Chakrabortty, R., Chandra Pal, S., Rezaie, F., Arabameri, A., Lee, S., Roy, P., Saha, A., Chowdhuri, I., & Moayedi, H. (2021). Flash-flood hazard susceptibility mapping in Kangsabati River Basin, India. Geocarto International, 37(23), 6713–6735. https://doi.org/10.1080/10106049.2021.1953618
10. Chen, T., & Guestrin, C. (2016). XGBoost: a scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794. https://doi.org/10.1145/2939672.2939785
11. Choi, J., Gu, B., Chin, S., & Lee, J. S. (2020). Machine learning predictive model based on national data for fatal accidents of construction workers. Automation in Construction, 110, Article 102974. https://doi.org/10.1016/j.autcon.2019.102974
12. Costache, R., Pham, Q. B., Avand, M., Thuy Linh, N. T., Vojtek, M., Vojteková, J., Lee, S., Khoi, D. N., Thao Nhi, P. T., & Dung, T. D. (2020). Novel hybrid models between bivariate statistics, artificial neural networks and boosting algorithms for flood susceptibility assessment. Journal of Environmental Management, 265, Article 110485. https://doi.org/10.1016/j.jenvman.2020.110485
13. Demir, V., & Ülke Keskin, A. (2022a). Taşkınların ekonomik zararlarının değerlendirilmesi (Samsun-Mert Irmağı havzası). Uluslararası Mühendislik Araştırma ve Geliştirme Dergisi, 14(2), 663–678. https://doi.org/10.29137/umagd.1090447
14. Demir, V., & Ülke Keskin, A. (2022b). Taşkın tehlike haritalarının oluşturulması (Samsun, Mert Irmağı örneği). Türkiye Coğrafi Bilgi Sistemleri Dergisi, 4(1), 47–54. https://doi.org/10.56130/tucbis.1120501
15. Demir, V., & Ülke Keskin, A. (2022c). Yeterince akım ölçümü olmayan nehirlerde taşkın debisinin hesaplanması ve taşkın modellemesi (Samsun, Mert Irmağı örneği). Geomatik, 7(2), 149–162. https://doi.org/10.29128/geomatik.918502
16. Deroliya, P., Ghosh, M., Mohanty, M. P., Ghosh, S., Rao, K. D., & Karmakar, S. (2022). A novel flood risk mapping approach with machine learning considering geomorphic and socio-economic vulnerability dimensions. Science of the Total Environment, 851, Article 158002. https://doi.org/10.1016/j.scitotenv.2022.158002
17. Devi, K. K., and Kumar, G. A. S. (2022). Stochastic gradient boosting model for twitter spam detection. Computer Systems Science and Engineering, 41 (2), 849–859. https://doi.org/10.32604/csse.2022.020836
18. Ekmekcioğlu, M., Başakın, E. E., & Özger, M. (2020). Tree-based nonlinear ensemble technique to predict energy dissipation in stepped spillways. European Journal of Environmental and Civil Engineering, 26(8), 3547–3565. https://doi.org/10.1080/19648189.2020.1805024
19. Ekmekcioğlu, M., & Koc, K. (2022a). Explainable step-wise binary classification for the susceptibility assessment of geo-hydrological hazards. Catena, 216, Article 106379. https://doi.org/10.1016/j.catena.2022.106379
20. Ekmekcioğlu, M., Koc, K., & Özger, M. (2022b). Towards flood risk mapping based on multi-tiered decision making in a densely urbanized metropolitan city of Istanbul. Sustainable Cities and Society, 80, Article 103759. https://doi.org/10.1016/j.scs.2022.103759
21. Ekmekcioğlu, M., Koc, K., Özger, M., & Işık, Z. (2022c). Exploring the additional value of class imbalance distributions on interpretable flash flood susceptibility prediction in the Black Warrior River basin, Alabama, United States. Journal of Hydrology, 610, Article 127877. https://doi.org/10.1016/j.jhydrol.2022.127877
22. ESRI. (2016, Haziran 30). What's new in ArcGIS 10.3.1 for Desktop. 30 Haziran 2016’da https://desktop.arcgis.com/en/arcmap/10.3/ main/get-started/whats-new-in-arcgis-1031.htm adresinden alındı.
23. Feng, Y., Wang, D., Yin, Y., Li, Z., & Hu, Z. (2020). An XGBoost-based casualty prediction method for terrorist attacks. Complex & Intelligent Systems, 6(3), 721–740. https://doi.org/10.1007/s40747-020-00173-0
24. Fiori, A., & Volpi, E. (2020). On the Effectiveness of LID Infrastructures for the attenuation of urban flooding at the catchment scale. Water Resources Research, 56(5), Article e2020WR027121. https://doi.org/10.1029/2020wr027121
25. Friedman, J. H. (2002). Stochastic gradient boosting. Computational Statistics & Data Analysis, 38(4), 367–378. https://doi.org/10.1016/s0167-9473(01)00065-2
26. Thippa Reddy G., Swarna Priya R.M., Parimala M., Chowdhary, C. L., M., P. K. R., Hakak, S., & Khan, W. Z. (2020). A deep neural networks based model for uninterrupted marine environment monitoring. Computer Communications, 157, 64–75. https://doi.org/10.1016/j.comcom.2020.04.004
27. Ghosh, A., Dey, P., & Ghosh, T. (2022). Integration of RS-GIS with Frequency Ratio, Fuzzy Logic, Logistic Regression and Decision Tree Models for Flood Susceptibility Prediction in Lower Gangetic Plain: A Study on Malda District of West Bengal, India. Journal of the Indian Society of Remote Sensing, 50(9), 1725–1745. https://doi.org/10.1007/s12524-022-01560-5
28. Guo, F., Zhou, X., Liu, J., Zhang, Y., Li, D., & Zhou, H. (2019). A reinforcement learning decision model for online process parameters optimization from offline data in injection molding. Applied Soft Computing, 85, Article 105828. https://doi.org/10.1016/j.asoc.2019.105828
29. Güvel, Ş. P., Akgül, M. A., & Akkoyunlu, M. F. (2023). Taşkın Riski Değerlendirmesi Kapsamında Landsat-8 Uydu Verileri ile 2015 Yılı Devrek Zonguldak Heyelanının İzlenmesi ve Değerlendirilmesi. Doğal Afetler ve Çevre Dergisi, 9(1), 81–89. https://doi.org/10.21324/dacd.1152670
30. Haltas, I., Yildirim, E., Oztas, F., & Demir, I. (2021). A comprehensive flood event specification and inventory: 1930–2020 Turkey case study. International Journal of Disaster Risk Reduction, 56, Article 102086. https://doi.org/10.1016/j.ijdrr.2021.102086
31. Hitouri, S., Varasano, A., Mohajane, M., Ijlil, S., Essahlaoui, N., Ali, S. A., Essahlaoui, A., Pham, Q. B., Waleed, M., Palateerdham, S. K., & Teodoro, A. C. (2022). Hybrid machine learning approach for gully erosion mapping susceptibility at a watershed scale. ISPRS International Journal of Geo-Information, 11(7), Article 401. https://doi.org/10.3390/ijgi11070401
32. Izonin, I., Tkachenko, R., Shakhovska, N., Ilchyshyn, B., & Singh, K. K. (2022). A two-step data normalization approach for improving classification accuracy in the medical diagnosis domain. Mathematics, 10(11), Article 1942. https://doi.org/10.3390/math10111942
33. Jaafar, H. H., Ahmad, F. A., & El Beyrouthy, N. (2019). GCN250, new global gridded curve numbers for hydrologic modeling and design. Scientific Data, 6(1), Article 145. https://doi.org/10.1038/s41597-019-0155-x
34. Kamalov, F. (2020). Kernel density estimation based sampling for imbalanced class distribution. Information Sciences, 512, 1192–1201. https://doi.org/10.1016/j.ins.2019.10.017
35. Kang, K., & Ryu, H. (2019). Predicting types of occupational accidents at construction sites in Korea using random forest model. Safety Science, 120, 226–236. https://doi.org/10.1016/j.ssci.2019.06.034
36. Kim, D., & Byun, J. (2022). Selection of augmented data for overcoming the imbalance problem in facies classification. IEEE Geoscience and Remote Sensing Letters, 19, 1–5. https://doi.org/10.1109/lgrs.2021.3103997
37. Koc, K., Ekmekcioğlu, M., & Gurgun, A. P. (2021). Integrating feature engineering, genetic algorithm and tree-based machine learning methods to predict the post-accident disability status of construction workers. Automation in Construction, 131, Article 103896. https://doi.org/10.1016/j.autcon.2021.103896
38. Koc, K., Ekmekcioğlu, M., & Gurgun, A. P. (2022). Prediction of construction accident outcomes based on an imbalanced dataset through integrated resampling techniques and machine learning methods. Engineering, Construction and Architectural Management, 30(9), 4486–4517. https://doi.org/10.1108/ecam-04-2022-0305
39. Kumar, S., Gupta, S., & Arora, S. (2022). A comparative simulation of normalization methods for machine learning-based intrusion detection systems using KDD Cup’99 dataset. Journal of Intelligent & Fuzzy Systems, 42(3), 1749–1766. https://doi.org/10.3233/jifs-211191
40. Liang, X., Jiang, A., Li, T., Xue, Y., & Wang, G. (2020). LR-SMOTE — An improved unbalanced data set oversampling based on K-means and SVM. Knowledge-Based Systems, 196, Article 105845. https://doi.org/10.1016/j.knosys.2020.105845
41. Meteoroloji Genel Müdürlüğü. (2023, Ocak 2023). Resmi İstatistikler. Meteoroloji Genel Müdürlüğü. 23 Ocak 2023’de https://www.mgm.gov.tr/veridegerlendirme/il-ve-ilceler-istatistik.aspx?k=H adresinden alındı.
42. Moisen, G. G., Freeman, E. A., Blackard, J. A., Frescino, T. S., Zimmermann, N. E., & Edwards, T. C. (2006). Predicting tree species presence and basal area in Utah: A comparison of stochastic gradient boosting, generalized additive models, and tree-based methods. Ecological Modelling, 199(2), 176–187. https://doi.org/10.1016/j.ecolmodel.2006.05.021
43. Nefeslioglu, H. A., Sezer, E., Gokceoglu, C., Bozkir, A. S., & Duman, T. Y. (2010). Assessment of landslide susceptibility by decision trees in the metropolitan area of Istanbul, Turkey. Mathematical Problems in Engineering, 2010, Article 901095. https://doi.org/10.1155/2010/901095
44. Norallahi, M., & Seyed Kaboli, H. (2021). Urban flood hazard mapping using machine learning models: GARP, RF, MaxEnt and NB. Natural Hazards, 106(1), 119–137. https://doi.org/10.1007/s11069-020-04453-3
45. Özcan, O. (2017). Taşkın tespitinin farklı yöntemlerle değerlendirilmesi: Ayamama deresi örneği. Doğal Afetler ve Çevre Dergisi, 3(1), 9–27. https://doi.org/10.21324/dacd.267200
46. Pham, Q. B., Ekmekcioğlu, M., Ali, S. A., Koc, K., & Parvin, F. (2023). Examining the role of class imbalance handling strategies in predicting earthquake-induced landslide-prone regions. Applied Soft Computing, 143, Article 110429. https://doi.org/10.1016/j.asoc.2023.110429
47. Provost, F. (2000). Machine learning from imbalanced data sets 101 [Invited paper]. AAAI'2000 Workshop on Imbalanced Data Sets, Menlo Park, CA.
48. Sarker, I. H., Alqahtani, H., Alsolami, F., Khan, A. I., Abushark, Y. B., & Siddiqui, M. K. (2020). Context pre-modeling: an empirical analysis for classification based user-centric context-aware predictive modeling. Journal of Big Data, 7(1), Article 51. https://doi.org/10.1186/s40537-020-00328-3
49. Şen, Z. (2009). İklim değişikliği içerikli taşkın afet ve modern hesaplama yöntemleri. Su Vakfı Yayınları.
50. Tehrany, M. S., Lee, M. J., Pradhan, B., Jebur, M. N., & Lee, S. (2014). Flood susceptibility mapping using integrated bivariate and multivariate statistical models. Environmental Earth Sciences, 72(10), 4001–4015. https://doi.org/10.1007/s12665-014-3289-3
51. Tien Bui, D., Hoang, N. D., Pham, T. D., Ngo, P. T. T., Hoa, P. V., Minh, N. Q., Tran, X. T., & Samui, P. (2019). A new intelligence approach based on GIS-based multivariate adaptive regression splines and metaheuristic optimization for predicting flash flood susceptible areas at high-frequency tropical typhoon area. Journal of Hydrology, 575, 314–326. https://doi.org/10.1016/j.jhydrol.2019.05.046
52. Tona, A. U., Demir, V., Kuşak, L., & Yakar, M. (2022). Su kaynakları mühendisliğinde CBS’nin kullanımı. Türkiye Coğrafi Bilgi Sistemleri Dergisi, 4(1), 23–33. https://doi.org/10.56130/tucbis.993807
53. Towfiqul Islam, A. R. M., Talukdar, S., Mahato, S., Kundu, S., Eibek, K. U., Pham, Q. B., Kuriqi, A., & Linh, N. T. T. (2021). Flood susceptibility modelling using advanced ensemble machine learning models. Geoscience Frontiers, 12(3), Article 101075. https://doi.org/10.1016/j.gsf.2020.09.006
54. Yu, R., & Abdel-Aty, M. (2014). Analyzing crash injury severity for a mountainous freeway incorporating real-time traffic and weather data. Safety Science, 63, 50–56. https://doi.org/10.1016/j.ssci.2013.10.012
55. Zhang, Z., Beck, M. W., Winkler, D. A., Huang, B., Sibanda, W., & Goyal, H. (2018). Opening the black box of neural networks: methods for interpreting neural network models in clinical applications. Annals of Translational Medicine, 6(11), 216–216. https://doi.org/10.21037/atm.2018.05.32