Automatic Detection of Earthquake Surface Ruptures with Deep Learning: The Case of the February 6, 2023 Kahramanmaraş Earthquakes

Öz

Earthquakes are a type of disaster that cause highest loss of life and property. Detecting and examining surface ruptures has great importance in understanding earthquakes and minimizing damage. This study aims to automatically detect surface ruptures that occurred during earthquakes with magnitudes of 7.7 and 7.6 that occurred in Kahramanmaraş province on February 6, 2023 and affected 11 provinces. MAXAR satellite images were used as the base image and YOLOv8n-seg was used as the segmentation model. An area where surface ruptures were visible was identified on the satellite images and the model was trained with the help of ruptures in this area. Of the images, 66% was used for training and 34% for testing. The mAP@0.5 value for the trained model was determined to be 0.88. The YOLOv8 model, as a method supported by deep learning techniques, offers high efficiency for these tasks. Using this method, surface ruptures that occur during earthquakes can be easily detected on satellite images. The method provides researchers with an important basis for field studies by allowing them to quickly detect surface fractures in large areas. In this way, time and economic losses will be minimized.

Anahtar Kelimeler

• Kahramanmaraş earthquakes
• Deep learning
• Segmentation
• YOLOv8
• Surface rupture

Derin Öğrenme ile Deprem Yüzey Kırıklarının Otomatik Belirlenmesi: 6 Şubat 2023 Kahramanmaraş Depremleri Örneği

Öz

Depremler dünyada en çok can ve mal kaybına neden olan afet türüdür. Bu yüzden deprem öncesi ve sonrası yapılacak çalışmalar önem kazanmaktadır. Depremler esnasında oluşan yüzey kırıklarının tespit edilmesi ve anlaşılması da depremlerin anlaşılmasında önem arz etmektedir. Bu çalışmada 02.06.2023 tarihinde Kahramanmaraş ilinde meydana gelen 7.7 ve 7.6 büyüklüğündeki depremlerde oluşmuş yüzey kırıklarının otomatik olarak tespit edilmesi amaçlanmıştır. Altlık olarak uydu görüntüleri, segmentasyon modeli olarak YOLOv8 kullanılmıştır. Uydu görüntüleri üzerinde yüzey kırıklarının gözle görülebildiği bir alan belirlenmiş ve bu alandaki kırıklar yardımı ile model eğitilmiştir. Verilerin %66 eğitim için %34 ise test için kullanılmıştır. Eğitilen modelin mAP@0.5 değeri 0.88 olarak tespit edilmiştir. YOLOv8 modeli, derin öğrenme teknikleriyle desteklenen bir yöntem olarak, bu tür görevlerde yüksek verimlilik vaat etmektedir. Bu yöntem kullanılarak deprem sonrası oluşacak yüzey kırıkları uydu görüntüleri üzerinde kolaylıkla tespit edilebilmektedir. Yöntem, araştırmacılara büyük alanlarda hızlı bir şekilde yüzey kırıklarının tespit edilmesine olanak sağlayarak arazi çalışmaları için önemli bir altlık sağlamaktadır. Bu sayede zaman ve ekonomik anlamdaki kayıplar en aza indirilmiş olacaktır.

Anahtar Kelimeler

• Kahramanmaraş depremi
• Derin öğrenme
• Segmentasyon
• YOLOv8
• Yüzey kırığı

Kaynakça

1. AFAD, (2023). 06 Şubat 2023 Pazarcık-Elbistan (Kahramanmaraş) Mw: 7.7 – Mw: 7.6 Depremleri Raporu. 140 s.
2. Akıncı, A. C., & Ünlügenç, U. C. (2023). 6 Şubat 2023 Kahramanmaraş Depremleri: Sahadan Jeolojik Veriler, Değerlendirme ve Adana için Etkileri. Çukurova Üniversitesi Mühendislik Fakültesi Dergisi, 38(2), 553-569. https://doi.org/10.21605/cukurovaumfd.1334155
3. Aksoy, E., Akgün, E., Softa, M., Koçbulut, F., vd. (2023). 6 Şubat 2023 Pazarcık (Kahramanmaraş) depreminin Doğu Anadolu Fay Zonu Erkenek ve Pazarcık Segmentleri Üzerindeki Etkisi: Çelikhan-Gölbaşı (Adıyaman) Arasından Gözlemler. Türk Deprem Araştırma Dergisi, 5(1), 85-104. https://doi.org/10.46464/tdad.1280408
4. Aydan, Ö. (2024). Crustal Stress State associated with the surface ruptures of the 2023 Pazarcık and Ekinözü Earthquakes. Yerbilimleri, 45(1), 52-65. https://doi.org/10.17824/yerbilimleri.1325646
5. Bayrakdar, C., Halis, O., Canpolat, E., Döker, M. F., vd. (2023). 6 Şubat 2023 Kahramanmaraş-Ekinözü depremi (Mw 7.6) ile ilişkili Çardak Fayı yüzey kırığının tektonik jeomorfolojisi. Türk Coğrafya Dergisi(83), 7-22. https://doi.org/10.17211/tcd.1281680
6. Brownlee, J. (2020). Data preparation for machine learning: Data cleaning, feature selection, and data transforms in Python. Machine Learning Mastery.
7. Chen, W., Zhong, C., Qin, X., & Wang, L. (2023). Deep learning based intelligent recognition of ground fissures. In Intelligent interpretation for geological disasters: from space-air-ground integration perspective (pp. 171-233). Singapore: Springer Nature Singapore. https://doi.org/10.1007/978-981-99-5822-1
8. Chen, X., Hu, G., & Liu, X. (2022). Recognition of earthquake surface ruptures using deep learning. Applied Sciences, 12(22), 11638. https://doi.org/10.3390/app122211638

9. Chen, Z. M., Wei, X. S., Wang, P., & Guo, Y. (2019). Multi-label image recognition with graph convolutional networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 5177-5186).
10. Choi, Y., Choi, J. H., & Choi, Y. (2023). Fit-for-purpose approach for the detection and analysis of earthquake surface ruptures using satellite images. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, 9574-9589. https://doi.org/10.1109/JSTARS.2023.3322347
11. Feng, J., Yan, J., & Zhao, X. (2023). Preliminary Investigation on the Surface Ruptures of the Turkey M7. 8 Earthquake on February 6, 2023. Acta Scientiarum Naturalium Universitatis Pekinensis, 59(6), 945-950.
12. García, S., Luengo, J., & Herrera, F. (2015). Data preprocessing in data mining. Springer. https://doi.org/10.1007/978-3-319-10247-4
13. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
14. Jafrasteh, B., Manighetti, I., & Zerubia, J. (2020). Generative adversarial networks as a novel approach for tectonic fault and fracture extraction in high resolution satellite and airborne optical images. In ISPRS International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences (Vol. 43, pp. 1219-1227). https://doi.org/10.5194/isprs-archives-XLIII-B3-2020-1219-2020
15. Kanakaraddi, S. G., Handur, V. S., Jalannavar, A., Chikaraddi, A., & Giraddi, S. (2024). Segmentation and Classification of Lung Cancer using Deep Learning Techniques. Procedia Computer Science, 235, 3226-3235. https://doi.org/10.1016/j.procs.2024.04.305
16. Kaneda, H., Nakata, T., Tsutsumi, H., Kondo, H., Sugito, N., Awata, Y., Akhtar, S.S., Majid, A., Khattak, W., Awan, A.A. and Yeats, R.S. (2008). Surface rupture of the 2005 Kashmir, Pakistan, earthquake and its active tectonic implications. Bulletin of the Seismological Society of America, 98(2), pp.521-557. https://doi.org/10.1785/0120070073
17. Kanoun, B., Cherif, M. A., Manighetti, I., Tarabalka, Y., & Zerubia, J. (2022). An enhanced deep learning approach for tectonic fault and fracture extraction in very high resolution optical images. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 3403-3407). IEEE. https://doi.org/10.1109/ICASSP43922.2022.9747007
18. Kelleher, J. D. (2023). Data science: An introduction. MIT Press.
19. Klinger, Y., Xu, X., Tapponnier, P., Van der Woerd, J., Lasserre, C., & King, G. (2005). High-resolution satellite imagery mapping of the surface rupture and slip distribution of the M w∼ 7.8, 14 November 2001 Kokoxili earthquake, Kunlun fault, northern Tibet, China. Bulletin of the Seismological Society of America, 95(5), 1970-1987. https://doi.org/10.1785/0120040233
20. Koçbulut, F., Softa, M., Akgün, E., Koşaroğlu, S., & Otlu, N. (2024) 6 Şubat 2023 (Mw: 7.6) Ekinözü Depremi Yüzey Kırığının Çığlık Fayı (Malatya-Doğanşehir-Eskiköy) Boyunca Karakteristik Analizi. Türk Deprem Araştırma Dergisi, 6(2), 387-404. https://doi.org/10.46464/tdad.1467421
21. Kotsiantis, S. B., Kanellopoulos, D., & Pintelas, P. E. (2006). Data preprocessing for supervised learning. International Journal of Computer Science, 1(2), 111–117.
22. Kozaci, O., Altunel, E., Koehler, R., Yildirim, C. and Clahan, K. (2024). M7. 8 Kahramanmaraş Earthquake Surface Fault Rupture and Near-Fault Effect Observations. Japanese Geotechnical Society Special Publication, 10(11), pp.276-281. https://doi.org/10.3208/jgssp.v10.SS-6-03
23. Kürçer, A., Özalp, S., Özdemir, E., Uygun Güldoğan, Ç., & Duman, T. Y. (2016). Yüzey Kırığı Oluşturmuş Faylar üzerinde aktif tektonik ve paleosismolojik araştırmalar hakkında örnek çalışma: Yenice Gönen Fayı, KB Türkiye. Maden Tetkik ve Arama Genel Müdürlüğü, Doğal Kay. ve Eko. Bült, 21, 1-18.
24. Li, K., Tapponnier, P., Xu, X., & Kang, W. (2023). The 2022, Ms 6.9 Menyuan earthquake: Surface rupture, Paleozoic suture re-activation, slip-rate and seismic gap along the Haiyuan fault system, NE Tibet. Earth and Planetary Science Letters, 622, 118412. https://doi.org/10.1016/j.epsl.2023.118412
25. Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014). Microsoft COCO: Common objects in context. In European Conference on Computer Vision (ECCV) (pp. 740–755). Springer. https://doi.org/10.1007/978-3-319-10602-1_48
26. Mattéo, L., Manighetti, I., Tarabalka, Y., Gaucel, J.M., van den Ende, M., Mercier, A., Tasar, O., Girard, N., Leclerc, F., Giampetro, T. and Dominguez, S. (2021). Automatic fault mapping in remote optical images and topographic data with deep learning. Journal of Geophysical Research: Solid Earth, 126(4), p.e2020JB021269. https://doi.org/10.1029/2020JB021269
27. Meng, J., Kusky, T., Mooney, W. D., Bozkurt, E., Bodur, M. N., & Wang, L. (2024). Surface deformations of the 6 February 2023 earthquake sequence, eastern Türkiye. Science, 383(6680), 298-305. https://doi.org/10.1126/science.adj3770
28. Mikhailov, V. O., Babayantz, I. P., Volkova, M. S., Timoshkina, E. P., Smirnov, V. B., & Tikhotskiy, S. A. (2023, July). The February 6, 2023, Earthquakes in Turkey: A Model of the Rupture Surface Based on Satellite Radar Interferometry. In Doklady Earth Sciences (Vol. 511, No. 1, pp. 571-577). Moscow: Pleiades Publishing. https://doi.org/10.1134/S1028334X2307019X
29. Patel, H., Shah, D., & Thakkar, A. (2024). Role of data preprocessing in modern machine learning pipelines. Journal of Big Data, 11, Article 25. https://doi.org/10.1186/s40537-024-00889-0
30. Patyniak, M., Landgraf, A., Dzhumabaeva, A., Baikulov, S., Williams, A.M., Preusser, F., Abdrakhmatov, K.E., Arrowsmith, J.R. and Strecker, M.R. (2024). Surface rupture of the 2008 Mw 6.6 Nura earthquake: Triggered flexural‐slip faulting in the Pamir‐Tien Shan collision zone. Tectonics, 43(9), p.e2024TC008360. https://doi.org/10.1029/2024TC008360
31. Polat, A., Keskin, İ., & Polat, Ö. (2023). Automatic Detection and Mapping of Dolines Using U-Net Model from Orthophoto Images. ISPRS International Journal of Geo-Information, 12(11), 456. https://doi.org/10.3390/ijgi12110456
32. Polat, A., Tatar, O., Gürsoy, H., Koçbulut, F., Mesci, B.L., Sezen, T.F., Kavak, K.Ş., Akpınar, Z., Töre, Y., Geyik, M. ve Yakan, M. (2006). Kuzey Anadolu Fay Zonu (KAFZ) doğusunda ÇobanlıAkçağıl (Suşehri-Sivas) arasında 1939 Erzincan Depremi yüzey kırığı. ATAG Aktif Tektonik Araştırma Grubu Toplantıları 10. Çalıştayı (2-4 Kasım 2006 İzmir) Bildiri Özleri Kitabı, s.72.
33. Polat, N. (2023). UAV-based investigation of earthquake-induced deformation and landscape changes: a case study of the February 6, 2023 Earthquakes in Hatay, Türkiye. Earth Science Informatics, 16(4), 3765-3777. https://doi.org/10.3390/s150715520
34. Polat, Ö., Polat, A., & Ekici, T. (2021). Automatic classification of volcanic rocks from thin section images using transfer learning networks. Neural computing and applications, 33(18), 11531-11540. https://doi.org/10.1007/s00521-021-05849-3
35. Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 779-788). https://doi.org/10.48550/arXiv.1506.02640
36. Reitman, N. G., Briggs, R. W., Barnhart, W. D., Hatem, A. E., Thompson Jobe, J. A., DuRoss, C. B., . & Akçiz, S. (2023). Rapid surface rupture mapping from satellite data: The 2023 Kahramanmaraş, Turkey (Türkiye), earthquake sequence. The Seismic Record, 3(4), 289-298. https://doi.org/10.1785/0320230029
37. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18 (pp. 234-241). Springer international publishing. https://doi.org/10.48550/arXiv.1505.04597
38. Softa, M. (2024). Determining the surface fault-rupture hazard zone for the Pazarcık segment of the East Anatolian fault zone through comprehensive analysis of surface rupture from the February 6, 2023, Earthquake (Mw 7.7). Journal of Mountain Science, 21(8), 2646-2663. https://doi.org/10.1007/s11629-024-8723-8
39. Tasar, O., Tarabalka, Y., & Alliez, P. (2019). Incremental learning for semantic segmentation of large-scale remote sensing data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(9), 3524-3537. https://doi.org/10.48550/arXiv.1810.12448
40. Ultralytics, (2024, 26 Şubat). Brief summary of YOLOv8 model structure #189. https://github.com/ultralytics/ultralytics/issues/189
41. Ünlügenç, U. C., Türkmen, S., Çetin, H., Güneyli, H., Nurlu, N., Akıncı, A.C. (2023). 06 ŞUBAT 2023 Kahramanmaraş Depremleri (Mw 7,8- Mw 7,6) Değerlendirme ve Saha Gözlem Raporu. Wang, L., Xia, Y., Zhang, Q., & Zhang, L. (2021). Annotation-efficient deep learning for automatic medical image segmentation. Computerized Medical Imaging and Graphics, 89, 101886. https://doi.org/10.1016/j.compmedimag.2021.101886
42. Wang, T., & Liu, S. (2023). Rapid surface rupture mapping from satellite data: The 2023 Turkey earthquakes. The Seismic Record, 3(4), 289–296. https://doi.org/10.1785/0320230023
43. Wang, Y., Wu, S.H., Chou, H.L.B., Li, Y.Y., Cheng, W.S., Ho, A., Chen, J.M., Liu, S.C., Hsieh, C.Y., Duan, S. and Min, S.M. (2024). Surface ruptures of the 2022 Guanshan-Chihshang earthquakes in central Longitudinal Valley area, eastern Taiwan. Terrestrial, Atmospheric and Oceanic Sciences, 35(1), p.16. https://doi.org/10.1016/j.soildyn.2021.106979
44. Yönlü, Ö., & Karabacak, V. (2024). Surface rupture history and 18 kyr long slip rate along the Pazarcık segment of the East Anatolian Fault. Journal of the Geological Society, 181(1), jgs2023-056. https://doi.org/10.1144/jgs2023-056
45. Yu, D., Ji, S., Li, X., Yuan, Z., & Shen, C. (2022). Earthquake crack detection from aerial images using a deformable convolutional neural network. IEEE Transactions on Geoscience and Remote Sensing, 60, 1-12. https://doi.org/10.1109/TGRS.2022.3183157
46. Yuan, Z., Liu-Zeng, J., Li, X., Xu, J., Yao, W., Han, L., & Li, T. (2021). Detailed mapping of the surface rupture of the 12 February 2014 Yutian M s 7.3 earthquake, Altyn Tagh fault, Xinjiang, China. Science China Earth Sciences, 64, 127-147. https://doi.org/10.1007/s11430-020-9673-6
47. Zeng, X., Liu, J., Wang, W., Yao, W., Wu, J., Liu, X., Han, L., Wang, W., Xing, Y., Du, R. and Yang, X. (2023). Machine learning in identifying and mapping the surface rupture of the 2021 M W 7.4 Madoi earthquake, Qinghai. Chinese Journal of Geophysics, 66(3), pp.1098-1112. https://doi.org/10.6038/cjg2022Q0117
48. Zhang, Y., & Yang, Q. (2017). A survey on multi-task learning. IEEE Transactions on Knowledge and Data Engineering, 29(12), 2027–2040. https://doi.org/10.1109/TKDE.2017.2699320
49. Zhang, Z., Wang, J., & Chen, X. (2023). Data quality management in AI applications: A survey. ACM Computing Surveys, 55(7), 1–36. https://doi.org/10.1145/3531090