Makine Öğrenmesi Algoritmalarının Türkiye Depremlerinde P ve S Fazlarının Belirlenmesindeki Etkinliği
Year 2025,
Volume: 7 Issue: 1, 90 - 100, 30.04.2025
Utku Ünal
,
Tolga Bekler
,
Feyza Nur Bekler
Abstract
Depremlerin dinamik parametrelerinin belirlenmesine yönelik analizlerde sismik cisim dalgalarının (P ve S fazları) varış zamanlarının doğru tespiti, sismolojik problemlerin çözümünde bir ön koşuldur ve varış zamanlarını temel alan çalışmalara, örneğin yer içi kabuk/manto yapısının anlaşılması gibi ve çeşitli araştırma alanlarına katkıda bulunur. Makine öğrenimi veya daha geniş ve bilinen uygulaması ile yapay zekâ teknolojilerindeki son gelişmeler, sismogramlar kullanılarak cisim dalgalarının varış zamanlarının otomatik olarak tespit edilmesini mümkün kılmaktadır. Türkiye, Avrasya, Afrika ve Arap levhalarının kesişim noktasında yer aldığı için karmaşık tektonik yapısı nedeniyle yüksek sismik aktiviteye sahiptir. Bu çalışmada, KRDAE (Kandilli Rasathanesi ve Deprem Araştırma Enstitüsü) ağına ait 2013-2019 yılları arasındaki Türkiye’de meydana gelen deprem kayıtlarında P ve S fazlarının geliş zamanlarını otomatik olarak tespit eden derin öğrenme algoritmasının başarımı değerlendirilmiştir. Sonuçlar, makine öğrenmesinin geleneksel istatistiksel yöntemlere kıyasla daha doğru tahminler yapabildiğini ve insan kaynaklı hataların azaltılmasında etkili olduğunu göstermektedir. Çalışmada elde edilen bulgular, büyük veri tabanlarıyla eğitilmiş derin öğrenme tabanlı sismik faz tespit algoritmalarının, yerel ihtiyaçlara uygun hale getirilmesiyle sismolojik analizlerde doğruluk ve hız artışı sağlayabileceğini ortaya koymaktadır. Gelecekteki çalışmalarda, yerel veriyle de eğitilmiş modellerin karşılaştırmalı olarak incelenmesi ve faz tespitinde uzman müdahalesine ihtiyaç duymayan algoritmaların geliştirilmesi önerilmektedir.
Supporting Institution
Finansal destek alınmamıştır.
Project Number
FBA-2024-4926
Thanks
Boğaziçi Üniversitesi Kandilli Rasathanesi ve Deprem Araştırma Enstitüsü Bölgesel Deprem-Tsunami İzleme Ve Değerlendirme Merkezi'ne veri kullanımını sağladığı için teşekkür ederiz.
Bu çalışmadaki hesaplama algoritmalarında kullanılan bilgisayar çevre donanımları, FBA-2024-4926 proje kodlu Çanakkale Onsekiz Mart Üniversitesi Bilimsel Araştırma Projeleri tarafından desteklenmiştir.
References
- Akaike H., 1985. Prediction and entropy, (In: A Celebration of Statistics, Editörler: Atkinson A. C., Fienberg S.E.,1985, Springer), 1–24 p.
- Allen C.R., 1982. Comparisons Between the North Anatolian Fault of Turkey and the San Andreas Fault of California, (In: Multidisciplinary Approach to Earthquake Prediction, Editörler: Işıkara A.M., Vogel A., Progress in Earthquake Prediction Research, vol 2, Vieweg+Teubner Verlag, Wiesbaden).
- Emre Ö., Duman T.Y., Özalp S., Elmacı H., Olgun Ş., Şaroğlu, F., 2013. Açıklamalı Türkiye Diri Fay Haritası, Ölçek 1:1.250.000. MTA Genel Müdürlüğü, Özel Yayın Serisi 30, Ankara.
- Mousavi S.M., Zhu W., Sheng Y., Beroza G.C., 2019. Cred: A deep residual network of convolutional and recurrent units for earthquake signal detection, Scientific Reports, 9, 10267.
- Münchmeyer J., Woollam J., Rietbrock A., Tilmann F., Lange D., Bornstein T., Diehl T., Giunchi C., Haslinger F., Jozinovic D., Michelini A., Saul J., Soto H., 2022. Which picker fits my data? A quantitative evaluation of deep learning based seismic pickers, Journal of Geophysical Research, 127(1), 1-22.
- NCEDC, 2014. Northern California Earthquake Data Center, UC Berkeley Seismological Laboratory Dataset.
- Perol T., Gharbi M., Denolle M., 2018. Convolutional neural network for earthquake detection and location, Science Advances, 4(2), e1700578.
- Powers D.M.W., 2011. Evaluation: From Precision, Recall and F-Measure to ROC, Informedness, Markedness & Correlation, Journal of Machine Learning Technologies, 2(1), 37-63.
- Ross Z.E., Meier M.A., Hauksson E., Heaton T.H., 2018. Generalized seismic phase detection with deep learning, Bulletin of the Seismological Society of America, 108(5A), 2894-2901.
- Soto H., Schurr B., 2020. DeepPhasePick: A method for detecting and picking seismic phases from local earthquakes based on highly optimized convolutional and recurrent deep neural networks, Geophysical Journal International, 227(2), 1268-1294.
- Unal U., Bekler T., 2024. Detection of P and S Wave Phases by Machine Learning using Northwestern Türkiye Local Seismic Network Data, Pure Appl. Geophys., 2024.
- Wessel P., Smith W.H.F., Scharroo R., Luis J., Wobbe F., 2013. GMT 5, Generic Mapping Tools: Improved Version Released, EOS, Transactions American Geophysical Union, 94(45), 409-410.
- Zhu W., Beroza G.C., 2019. PhaseNet: a deep-neural-network-based seismic arrival time picking method, Geophysical Journal International, 216(1), 261-273.
The Effectiveness of Machine Learning Algorithms in Identifying P and S Phases of Earthquakes in Türkiye
Year 2025,
Volume: 7 Issue: 1, 90 - 100, 30.04.2025
Utku Ünal
,
Tolga Bekler
,
Feyza Nur Bekler
Abstract
In analyses aimed at determining the dynamic parameters of earthquakes, the accurate detection of the arrival times of primary seismic waves (P and S waves) is a fundamental prerequisite for solving seismological problems. Studies based on these arrival times contribute to various research areas, such as understanding the Earth's crust and mantle structure. Recent developments in machine learning—or more broadly and commonly known as artificial intelligence technologies—have made it possible to automatically detect earthquake waves from waveform data. Due to its complex tectonic structure, Türkiye has high seismic activity, as it is located at the intersection point of the Eurasian, African, and Arabian plates. In this study, the performance of a deep learning algorithm that automatically detects the arrival times of P and S phases in earthquakes occurred in Türkiye from 2013 to 2019 belonging to KOERI (Kandilli Observatory and Earthquake Research Institute) network was evaluated. The results show that machine learning can make more accurate predictions compared to traditional statistical methods and is effective in reducing human-induced errors. The findings of the study reveal that deep learning-based seismic phase detection algorithms trained with large databases can provide increased accuracy and speed in seismological analyses when adapted to local needs. It is recommended that future studies conduct comparative examinations of models also trained with local data and develop algorithms that do not require human intervention in phase detection.
Project Number
FBA-2024-4926
References
- Akaike H., 1985. Prediction and entropy, (In: A Celebration of Statistics, Editörler: Atkinson A. C., Fienberg S.E.,1985, Springer), 1–24 p.
- Allen C.R., 1982. Comparisons Between the North Anatolian Fault of Turkey and the San Andreas Fault of California, (In: Multidisciplinary Approach to Earthquake Prediction, Editörler: Işıkara A.M., Vogel A., Progress in Earthquake Prediction Research, vol 2, Vieweg+Teubner Verlag, Wiesbaden).
- Emre Ö., Duman T.Y., Özalp S., Elmacı H., Olgun Ş., Şaroğlu, F., 2013. Açıklamalı Türkiye Diri Fay Haritası, Ölçek 1:1.250.000. MTA Genel Müdürlüğü, Özel Yayın Serisi 30, Ankara.
- Mousavi S.M., Zhu W., Sheng Y., Beroza G.C., 2019. Cred: A deep residual network of convolutional and recurrent units for earthquake signal detection, Scientific Reports, 9, 10267.
- Münchmeyer J., Woollam J., Rietbrock A., Tilmann F., Lange D., Bornstein T., Diehl T., Giunchi C., Haslinger F., Jozinovic D., Michelini A., Saul J., Soto H., 2022. Which picker fits my data? A quantitative evaluation of deep learning based seismic pickers, Journal of Geophysical Research, 127(1), 1-22.
- NCEDC, 2014. Northern California Earthquake Data Center, UC Berkeley Seismological Laboratory Dataset.
- Perol T., Gharbi M., Denolle M., 2018. Convolutional neural network for earthquake detection and location, Science Advances, 4(2), e1700578.
- Powers D.M.W., 2011. Evaluation: From Precision, Recall and F-Measure to ROC, Informedness, Markedness & Correlation, Journal of Machine Learning Technologies, 2(1), 37-63.
- Ross Z.E., Meier M.A., Hauksson E., Heaton T.H., 2018. Generalized seismic phase detection with deep learning, Bulletin of the Seismological Society of America, 108(5A), 2894-2901.
- Soto H., Schurr B., 2020. DeepPhasePick: A method for detecting and picking seismic phases from local earthquakes based on highly optimized convolutional and recurrent deep neural networks, Geophysical Journal International, 227(2), 1268-1294.
- Unal U., Bekler T., 2024. Detection of P and S Wave Phases by Machine Learning using Northwestern Türkiye Local Seismic Network Data, Pure Appl. Geophys., 2024.
- Wessel P., Smith W.H.F., Scharroo R., Luis J., Wobbe F., 2013. GMT 5, Generic Mapping Tools: Improved Version Released, EOS, Transactions American Geophysical Union, 94(45), 409-410.
- Zhu W., Beroza G.C., 2019. PhaseNet: a deep-neural-network-based seismic arrival time picking method, Geophysical Journal International, 216(1), 261-273.