Multiple Regression-Based Prediction Method to Assess the Impact of PGA and Distance on Post-Earthquake Structural Damage Levels

Abstract

This study was carried out to evaluate the accuracy of the damage assessments made after the 06 February 2023 Kahramanmaraş earthquakes and to ensure that these data are a guide for future studies in the field of earthquake engineering. The relationship between damage levels, maximum ground acceleration (PGA) values measured by Disaster and Emergency Management Affair (DEMA) stations and distances to earthquake-affected cities were analyzed. Unlike the studies in literature, evaluation was made on multiple input and multiple output parameters, and a separate regression model was used for each output data. As a result of regression analysis, a significant relationship was found between damage levels and PGA-distance parameters. The R² scores for the "No damage" and "Heavy damage" levels were found to be 0.75 and 0.71, respectively. In the analyzes made by reducing the damage levels to two main categories (damaged and undamaged), the R² scores were calculated as 0.63 and 0.6, respectively. These results show that there is a sufficient level of agreement between the input and output parameters, but they reveal that the dataset should be expanded, and the positional details of the structures should be obtained separately for higher accuracy. Within the scope of the study, linear regression, polynomial regression, random forest and gradient boosting models were used and their performances were compared. According to the results obtained, gradient boosting and random forest models were the models that exhibited the best compatibility according to damage levels. In particular, the fact that the random forest model gives the best results in 5 out of 6 damage levels shows that the model is a method that produces fast and reliable results in such complex analyses. As a result, it was determined that model performance at low conforming damage levels could be improved by expanding the data set and increasing the available data details. These findings make important contributions to the accuracy analysis of damage assessments after earthquakes and provide a scientific basis for similar studies.

Keywords

• Multiple Regression
• Artificial Intelligence
• Damage Distribution
• Earthquake Observation
• Kahramanmaras Earthquakes

Multiple Regression-Based Prediction Method to Assess the Impact of PGA and Distance on Post-Earthquake Structural Damage Levels

Abstract

Bu çalışma, 06 Şubat Kahramanmaraş depremleri sonrasında yapılan hasar tespitlerinin doğruluğunu değerlendirmek ve bu verilerin deprem mühendisliği alanındaki gelecekteki çalışmalar için rehber olmasını sağlamak amacıyla gerçekleştirilmiştir. Hasar seviyeleri, Afet ve Acil Durum Yönetimi Başkanlığı (AFAD) istasyonlarının ölçtüğü en büyük yer ivmesi (PGA) değerleri ve depremden etkilenen şehirlere olan mesafeleri arasındaki ilişki analiz edilmiştir. Literatürdeki çalışmalardan farklı olarak, çoklu giriş ve çoklu çıkış parametreleri üzerinden değerlendirme yapılmış ve her bir çıkış verisi için ayrı regresyon modeli kullanılmıştır. Regresyon analizleri sonucunda, hasar seviyeleri ile PGA-mesafe parametreleri arasında anlamlı bir ilişki tespit edilmiştir. "Hasar yok" ve "Ağır hasar" seviyeleri için R² skorları sırasıyla 0.75 ve 0.71 olarak bulunmuştur. Hasar seviyeleri iki ana kategoriye (hasarlı ve hasarsız) indirgenerek yapılan analizlerde ise R² skorları sırasıyla 0.63 ve 0.6 olarak hesaplanmıştır. Bu sonuçlar, giriş ve çıkış parametreleri arasında yeterli düzeyde uyum olduğunu göstermekle birlikte, daha yüksek doğruluk için veri setinin genişletilmesi ve yapıların konumsal detaylarının ayrı ayrı elde edilmesi gerektiğini ortaya koymaktadır. Çalışma kapsamında lineer regresyon, polinomal regresyon, random forest ve gradient boosting modelleri kullanılmış ve performansları karşılaştırılmıştır. Elde edilen sonuçlara göre gradient boosting ve random forest modelleri, hasar seviyelerine göre en iyi uyumu sergileyen modeller olmuştur. Özellikle random forest modelinin 6 hasar seviyesinden 5’inde en iyi sonuçları vermesi, bu tür karmaşık analizlerde modelin hızlı ve güvenilir sonuçlar üreten bir yöntem olduğunu göstermektedir. Sonuç olarak, düşük uyum gösteren hasar seviyelerinde model performansının, veri setinin genişletilmesi ve mevcut veri detaylarının artırılmasıyla iyileştirilebileceği belirlenmiştir. Bu bulgular, depremler sonrası hasar tespitlerinin doğruluk analizine önemli katkılar sağlamakta ve benzer çalışmalar için bilimsel bir temel oluşturmaktadır.

Keywords

• Çoklu Regresyon
• Yapay Zeka
• Hasar Dağılımı
• Deprem Gözlemi
• Kahramanmaraş Depremleri

References

1. Nemutlu ÖF, Balun B, Sari A. Damage assessment of buildings after 24 January 2020 Elazig- Sivrice earthquake. Earthquakes and Structures. 25 Mart 2021;20(3):325-35.
2. Balun B, Nemutlu OF, Benli A, Sari A. Estimation of probabilistic hazard for Bingol province, Turkey. Earthquakes and Structures. 25 Şubat 2020;18(2):223-31.
3. Avcil F, Işık E, İzol R, Büyüksaraç A, Arkan E, Arslan MH, vd. Effects of the February 6, 2023, Kahramanmaraş earthquake on structures in Kahramanmaraş city. Nat Hazards. Şubat 2024;120(3):2953-91.
4. Altunsu E, Güneş O, Öztürk S, Sorosh S, Sarı A, Beeson ST. Investigating the structural damage in Hatay province after Kahramanmaraş-Türkiye earthquake sequences. Engineering Failure Analysis. 2024;157:107857.
5. Öztürk S, Altunsu E, Güneş O, Sarı A. Investigation of industrial structure performances in the Hatay and Gaziantep provinces during the Türkiye earthquakes on February 6, 2023. Soil Dynamics and Earthquake Engineering. Nisan 2024;179:108569.
6. Nemutlu ÖF, Balun B, Sarı A. 06 Şubat 2023 Kahramanmaraş Depremleri Kaynaklı Yapısal Hasarların Adıyaman İli Özelinde İncelenmesi. İçinde Konya; 2023.
7. Hussain E, Kalaycıoğlu S, Milliner CW, Çakir Z. Preconditioning the 2023 Kahramanmaraş (Türkiye) earthquake disaster. Nature Reviews Earth & Environment. 2023;4(5):287-9.
8. AFAD. 06 Şubat Kahramanmaraş(Pazarcık ve Elbistan) Depremleri Saha Çalışmaları Ön Değerlendirme Raporu. AFAD; 2023.

9. Şenol Balaban M, Doğulu C, Akdede N, Akoğlu H, Karakayalı O, Yılmaz S, vd. Emergency response, and community impact after February 6, 2023 Kahramanmaraş Pazarcık and Elbistan Earthquakes: reconnaissance findings and observations on affected region in Türkiye. Bulletin of Earthquake Engineering. 01 Şubat 2025;23(3):1053-81.
10. Altunişik AC, Arslan ME, Kahya V, Aslan B, Sezdirmez T, Dok G, vd. Field Observations and Damage Evaluation in Reinforced Concrete Buildings After the February 6th, 2023, Kahramanmaraş–Türkiye Earthquakes. J Earthquake and Tsunami. 01 Aralık 2023;17(06):2350024.
11. Yuzbasi J. Post-Earthquake Damage Assessment: Field Observations and Recent Developments with Recommendations from the Kahramanmaraş Earthquakes in Türkiye on February 6th, 2023 (Pazarcık M7.8 and Elbistan M7.6). Journal of Earthquake Engineering. :1-26.
12. Avğın S, Köse MM, Özbek A. Damage assessment of structural and geotechnical damages in Kahramanmaraş during the February 6, 2023 earthquakes. Engineering Science and Technology, an International Journal. 01 Eylül 2024;57:101811.
13. Işık E, Hadzima-Nyarko M, Avcil F, Büyüksaraç A, Arkan E, Alkan H, vd. Comparison of Seismic and Structural Parameters of Settlements in the East Anatolian Fault Zone in Light of the 6 February Kahramanmaraş Earthquakes. Infrastructures. 2024;9(12).
14. Tao D, Cai Y. Study on the Relation Between Ground Motion Parameters and Simulated Earthquake Damage of Simplified Masonry Structures. 2018.
15. Zhou Q, Sun B. Study on Earthquake Damage Distribution of Multistory Masonry Buildings. The Open Civil Engineering Journal. 2015;9(1):435-41.
16. Liang H, Li J, Yang Z, Li YQ. Effects of Epicentral Distance and Seismogenic Fault Distance on Seismic Damage of Dams in Wenchuan Earthquake. Applied Mechanics and Materials. 2013;353-356:2187-90.
17. Karaşin İB. Comparative Analysis of the 2023 Pazarcık and Elbistan Earthquakes in Diyarbakır. Buildings. 2023;13(10):2474.
18. Zengin B, Aydin F. The Effect of Material Quality on Buildings Moderately and Heavily Damaged by the Kahramanmaraş Earthquakes. Applied Sciences. 2023;13(19):10668.
19. Choi H, Sanada Y, KASHIWA H, Watanabe Y, Tanjung J, Jiang H. Seismic Response Estimation Method for Earthquake‐damaged RC buildings. Earthquake Engineering & Structural Dynamics. 2016;45(6):999-1018.
20. Huang HC. Characteristics of earthquake ground motions and the H/V of microtremors in the southwestern part of Taiwan. Earthquake Engineering and Structural Dynamics. 2002;31(10):1815-29.
21. Bilen A, Özer AB. Regresyon Yöntemlerine Dayalı Suç Tespit Analizi Karşılaştırması Elazığ İli Örneği. Fırat Üniversitesi Mühendislik Bilimleri Dergisi. 2022;34(1):115-21.
22. Shen Y, Wang L, Jian W, Shang J, Wang X, Ju L, vd. Big-Data and Artificial-Intelligence-Assisted Vault Prediction and EVO-ICL Size Selection for Myopia Correction. British Journal of Ophthalmology. 2021;107(2):201-6.
23. Suparwito H, Polina AM. Prediction of Tobacco Leave Grades With Ensemble Machine Learning Methods. 2019;1-6.
24. Liou L, Mostofsky E, Lehman LL, Salia S, Barrera FJ, Ying W, vd. Survival Machine Learning Methods for Mortality Prediction After Heart Transplantation in the Contemporary Era. Plos One. 2025;20(1):e0313600.
25. Pahno S, Yang J, Kim SS. Use of Machine Learning Algorithms to Predict Subgrade Resilient Modulus. Infrastructures. 2021;6(6):78.
26. Özçelik STA, Üzen H, Şengür A, Çelebi A. Derin öğrenme ile panoramik radyografi görüntülerinden diş segmentasyonu: UNet, FPN ve PSPNet karşılaştırması. Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi. 2024;13(4):1347-54.
27. Yusufoğlu E, Fırat H, Üzen H, Özçelik STA, Çiçek İB, Şengür A, vd. A Comprehensive CNN Model for Age-Related Macular Degeneration Classification Using OCT: Integrating Inception Modules, SE Blocks, and ConvMixer. Diagnostics. 2024;14(24):2836.
28. Üzen H. İmalat Sistemlerinde Derin Öğrenme Tabanlı Doku Hata Tespiti [Doktora]. [Fen Bilimleri Enstitüsü]: İnönü Üniversitesi; 2022.
29. Mangalathu S, Sun H, Nweke CC, Yi Z, Burton HV. Classifying earthquake damage to buildings using machine learning. Earthquake Spectra. 01 Şubat 2020;36(1):183-208.
30. Xia H, Wu J, Yao J, Zhu H, Gong A, Yang J, vd. A Deep Learning Application for Building Damage Assessment Using Ultra-High-Resolution Remote Sensing Imagery in Turkey Earthquake. International Journal of Disaster Risk Science. 01 Aralık 2023;14(6):947-62.
31. Nemutlu ÖF. Detection of earthquake damage using pre and past-earthquake satellite data. İçinde Tashkent, Uzbekistan; 2024. s. 256-66.
32. Aloisio A, Rosso MM, De Leo AM, Fragiacomo M, Basi M. Damage classification after the 2009 L’Aquila earthquake using multinomial logistic regression and neural networks. International Journal of Disaster Risk Reduction. Ekim 2023;96:103959.
33. Sheibani M, Ou G. The development of Gaussian process regression for effective regional post-earthquake building damage inference. Computer-Aided Civil and Infrastructure Engineering. 01 Mart 2021;36(3):264-88.
34. Adanur S, Altunişik AC, Bayraktar A, Akköse M. Comparison of near-fault and far-fault ground motion effects on geometrically nonlinear earthquake behavior of suspension bridges. Natural Hazards. 2012;64(1):593-614.
35. Işık E, Avcil F, Büyüksaraç A, İzol R, Arslan MH, Aksoylu C, vd. Structural damages in masonry buildings in Adıyaman during the Kahramanmaraş (Turkiye) earthquakes (Mw 7.7 and Mw 7.6) on 06 February 2023. Engineering Failure Analysis. 2023;151(May).
36. Nemutlu ÖF, Sari A, Balun B. 06 Şubat 2023 Kahramanmaraş Depremlerinde (Mw 7.7-Mw 7.6) Meydana Gelen Gerçek Can Kayıpları Ve Yapısal Hasar Değerlerinin Tahmin Edilen Değerler İle Karşılaştırılması. Afyon Kocatepe University Journal of Sciences and Engineering. 27 Ekim 2023;23(5):1222-34.
37. AFAD. 06 Şubat 2023 Pazarcık-Elibstan Kahramanmaraş(Mw:7.7-Mw:7.6) Depremleri Raporu. AFAD: AFAD; 2023 Haz s. 140.
38. AFAD. tadas.afad.gov.tr. 2023.
39. Mahesh B. Machine Learning Algorithms -A Review. C. 9, International Journal of Science and Research (IJSR). 2019.
40. Sarker IH. Machine Learning: Algorithms, Real-World Applications and Research Directions. SN Computer Science. 22 Mart 2021;2(3):160.
41. Alzubaidi L, Zhang J, Humaidi AJ, Al-Dujaili A, Duan Y, Al-Shamma O, vd. Review of deep learning: concepts, CNN architectures, challenges, applications, future directions. Journal of Big Data. 31 Mart 2021;8(1):53.
42. Mishra C, Gupta D. Deep Machine Learning and Neural Networks: An Overview. IAES International Journal of Artificial Intelligence (IJ-AI). 01 Haziran 2017;6:66.
43. Ostertagova E. Modelling Using Polynomial Regression. Procedia Engineering. 31 Aralık 2012;48:500-6.
44. Breiman L. Random Forests. Machine Learning. 01 Ekim 2001;45(1):5-32.
45. Zemel RS, Pitassi T. A Gradient-Based Boosting Algorithm for Regression Problems.
46. Çevre Şehircilik ve İklim Değişikliği Bakanlığı. csb.gov.tr. Basın Bülteni. 2023.
47. Python [Internet]. Python.org. 2024 [a.yer 21 Aralık 2024]. Erişim adresi: https://www.python.org/
48. Nadkarni SB, Vijay GS, Kamath RC. Comparative Study of Random Forest and Gradient Boosting Algorithms to Predict Airfoil Self-Noise. Engineering Proceedings. 2023;59(1).
49. Sousa M, Sant’Ana R, Fernandes R, Duarte J, Aploinário J, Thomä R. Comparison of Random Forest and Gradient Boosting Fingerprints to Enhance an Outdoor Radio-frequency Localization System. 2020.