Forecasting Earthquake Impact Scenarios in Istanbul with Machine Learning Algorithms

Yıl 2024, Cilt: 14 Sayı: 3, 95 - 105, 25.11.2024

Remzi Gürfidan , Mehmet Ali Yalçınkaya

Öz

This study employs machine learning algorithms to forecast the impacts of a potential magnitude 7.5 earthquake in Istanbul, focusing on casualty rates, hospitalization needs, and temporary shelter requirements. Using a dataset compiled from the Istanbul Metropolitan Municipality Open Data Portal and the Turkish Statistical Institute, the research assesses Gradient Boosting, AdaBoost, Random Forest, and ExtraTrees algorithms. Gradient Boosting emerged as the most effective model, exhibiting high accuracy and low prediction errors in determining disaster impacts. This approach underscores the critical role of advanced analytics in enhancing urban disaster preparedness and management, providing valuable insights for policymaking and infrastructure development in earthquake-prone areas.

Anahtar Kelimeler

Earthquake impact forecasting, disaster preparedness, machine learning, urban risk management

Proje Numarası

-

İstanbul’daki Deprem Etki Senaryolarının Makine Öğrenmesi Algoritmaları ile Tahmin Edilmesi

Öz

Bu çalışmada, İstanbul’da olası bir 7.5 büyüklüğündeki depremin etkilerini, özellikle de can kaybı sayısı, hastaneye ihtiyaç duyacak kişi sayısı ve geçici barınma ihtiyacı duyacak kişi sayısını tahmin etmek için makine öğrenmesi algoritmaları kullanılmaktadır. İstanbul Büyükşehir Belediyesi Açık Veri Portalı ve Türkiye İstatistik Kurumu’ndan derlenen bir veri seti kullanılarak Gradyan Artırma (Gradient Boosting), Uyarlanabilir Artırma (AdaBoost), Rastgele Orman (Random Forest) ve Ekstra Ağaçlar (ExtraTrees) algoritmaları değerlendirilmiştir. Gradient Boosting modeli, yüksek doğruluk ve düşük tahmin hataları ile en etkili model olarak öne çıkmıştır. Bu yaklaşım, gelişmiş analitiklerin kentsel afet hazırlığı ve yönetimini geliştirme konusundaki kritik rolünü vurgulamakta ve depreme eğilimli bölgelerdeki alınacak önlemler ve altyapı gelişimi için değerli öngörüler sağlamaktadır.

Anahtar Kelimeler

Deprem etki tahmini, afet hazırlığı, makine öğrenmesi, kentsel risk yönetimi

Etik Beyan

-

Destekleyen Kurum

-

Proje Numarası

-

Teşekkür

-

Kaynakça

• AlOmar, M.K., Hameed, M.M., AlSaadi M.A., 2020. Multi hours ahead prediction of surface ozone gas concentration. Robust artificial intelligence approach Atmospheric Pollution Research, 11(9): 1572–1587. Doi: 10.1016/j.apr.2020.06.024
• Ahamed, S., Daub, E. 2019. Machine learning approach to earthquake rupture dynamics. ArXiv, Doi:10.48550/arXiv.1906.06250
• Ambraseys, N., 2001. The earthquake of 10 July 1894 in the Gulf of Izmit (Turkey) and its relation to the earthquake of 17 August 1999. Journal of Seismology, 5(1): 117-128. Doi: 10.1023/A:1009871605267
• Arslan, G., Fawzy, D., Coskun, A. 2017. On the prediction of structural reactions to big earthquakes in Turkey. PressAcademia Procedia. 5(1): 335-340. Doi: 10.17261/Pressacademia.2017.608
• Breiman, L., 2001. Random forests. Machine learning. 45(1): 5–32. Doi: 10.1023/A:1010933404324
• Corbi, F., Sandri, L., Bedford, J., Funiciello, F., Brizzi, S., Rosenau, M., Lallemand, S. 2019. Machine learning can predict the timing and size of analog earthquakes. Geophysical Research Letters, 46: 1303 - 1311. Doi:10.1029/2018GL081251.
• Erdik, M. 2001. Report on 1999 Kocaeli and Düzce (Turkey) earthquakes. Structural Control For Civil And Infrastructure Engineering. p. 149-186. Paris. France.
• Freund, Y., Schapire, R.E. 1997. A decision-theoretic generalization of on-line learning and an application to boosting. Journal of Computer and System Sciences. 55(1): 119-139. Doi: 10.1006/jcss.1997.1504
• Friedman, J.H. 2001. Greedy function approximation: A gradient boosting machine. Annals of Statistics. 29(5):1189-1232. Doi: 10.1214/aos/1013203451
• Geurts, P., Ernst, D., Wehenkel, L., 2006. Extremely randomized trees. Machine Learning. 63(1): 3–42. Doi: 10.1007/s10994-006-6226-1.
• Hammid, A.T., Sulaiman, M.H, Abdalla, A.N. 2018. Prediction of small hydropower plant power production in Himreen Lake dam (HLD) using artificial neural network. Alexandria Engineering Journal. 57(1): 211–221. Doi: 10.1016/j.aej.2016.12.011.
• IMM Open Data Portal 2017. Neighborhood based building numbers. https://data.ibb.gov.tr/dataset/mahalle-bazli-bina-analiz-verisi/resource/cef193d5-0bd2-4e8d-8a69-275c50288875
• IMM Open Data Portal 2021. Earthquake scenario analysis results. https://data.ibb.gov.tr/dataset/deprem-senaryosu-analiz-sonuclari/resource/9c3ac492-de4b-4245-b418-7ad3df67a193
• James, G., Witten, D., Hastie, T., Tibshirani, R. 2013. An introduction to statistical learning. Springer. New York, USA, 607 pp.
• Jaiswal, K., Wald, D.J. 2010. An empirical model for global earthquake fatality estimation. Earthquake Spectra, 26(4): 1017-1037. Doi: 10.1193/1.3480331
• Korkut, U., Basbugoglu, T., Sahin, O. 2023. Turkey’s 2023 elections: another victory for Erdogan. Political Insight, 14(3): 16-19. Doi: 10.1177/20419058231198579
• Li, B., Gong, A., Zeng, T., Bao, W., Xu, C., Huang, Z. 2021. A zoning earthquake casualty prediction model based on machine learning. Remote Sensing, 14(1): 30. Doi: 10.3390/rs14010030
• Mishra, G., Sehgal, D, Valadi, J., K. 2017. Quantitative structure activity relationship study of the antihepatitis peptides employing random forests and extra-trees regressors. Bioinformation, 13(3): 60, Doi: 10.6026/97320630013060.
• Parsons, T., 2000. Heightened odds of large earthquakes near Istanbul: An interaction-based probability calculation. Science, 288(5466): 661-665, Doi: 10.1126/science.288.5466.661
• Rouet-Leduc, B., Hulbert, C., Lubbers, N., Barros, K., Humphreys, C., Johnson, P. 2017. Machine learning predicts laboratory earthquakes. Geophysical Research Letters. 44: 9276 - 9282. Doi:10.1002/2017GL074677.
• Stein, R.S. 2000. Earthquake conversations. Scientific American, 282(1): 68-75. Doi:10.1038/scientificamerican012003-DipeiQnMaPMOL1RMjOPhi
• Willmott, C.J., Matsuura, K. 2005. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Climate Research. 30(1): 79–82, Doi: 10.3354/cr030079.
• World Bank 2023. Earthquake damage in Türkiye estimated to exceed $34 billion: World bank disaster assessment report. https://www.worldbank.org
• Xing, H., Junyi, S., Jin, H. 2020. The casualty prediction of earthquake disaster based on Extreme Learning Machine method. Natural Hazards, 102(3): 873-886, Doi: 10.1007/s11069-020-03937-6