Quantifying Post-Earthquake Urban Green Space Loss Using an Optimized Deep Learning-Based Semantic Segmentation Framework

Öz

Large-scale earthquakes and the rapid reconstruction processes that follow lead to significant spatial transformations in urban ecosystems, particularly exerting pronounced impacts on intra-urban green spaces. This study quantitatively evaluates changes in urban green areas that occurred after the 2023 Kahramanmaraş earthquakes in Hatay (Antakya) and Elazığ (Güneykent) by utilizing high-resolution satellite imagery and an optimized deep learning–based semantic segmentation approach. In the proposed framework, an enhanced U-Net architecture incorporating asymmetric and dilated convolutional layers was designed to improve parameter efficiency while preserving extensive contextual information. The proposed model, comprising approximately 1.7 million parameters, achieved 88% accuracy, 0.76 Intersection over Union (IoU), and a 0.86 Dice score, demonstrating competitive performance compared to contemporary segmentation models with substantially higher parameter counts. Pixel-level comparison of pre- and post-earthquake imagery revealed reductions in green space reaching up to 89% in the examined sub-regions, with even higher rates observed in certain localized areas. The findings indicate that intensive post-disaster construction and urban reorganization processes may impose substantial pressure on environmental sustainability. The proposed method provides a computationally efficient and scalable decision-support framework for monitoring post-disaster environmental changes and developing sustainable urban planning strategies.

Anahtar Kelimeler

• Pixel-level analysis
• Optimized U-Net architecture
• Comparative performance evaluation
• Spatial sustainability impact

Optimize Edilmiş Derin Öğrenme Tabanlı Semantik Bölümleme Çerçevesi Kullanarak Deprem Sonrası Kentsel Yeşil Alan Kaybının Nicel Olarak Belirlenmesi

Öz

Büyük ölçekli depremler sonrasında gerçekleştirilen hızlı yeniden yapılanma süreçleri, kentsel ekosistemlerde önemli mekânsal dönüşümlere yol açmakta ve özellikle kent içi yeşil alanlar üzerinde belirgin etkiler oluşturmaktadır. Bu çalışma, 2023 Kahramanmaraş depremleri sonrasında Hatay (Antakya) ve Elazığ (Güneykent) bölgelerinde meydana gelen kentsel yeşil alan değişimlerini yüksek çözünürlüklü uydu görüntüleri ve optimize edilmiş derin öğrenme tabanlı anlamsal segmentasyon yaklaşımı kullanarak nicel olarak değerlendirmektedir. Çalışmada, parametre verimliliğini artırmak ve geniş bağlamsal bilgiyi korumak amacıyla asimetrik ve genişletilmiş (dilated) evrişim katmanları içeren geliştirilmiş bir U-Net mimarisi tasarlanmıştır. Yaklaşık 1.7 milyon parametreye sahip olan önerilen model, %88 doğruluk, 0.76 IoU ve 0.86 Dice skoru elde etmiş; daha yüksek parametre sayısına sahip çağdaş segmentasyon modelleriyle karşılaştırıldığında rekabetçi performans göstermiştir. Deprem öncesi ve sonrası görüntülerin piksel düzeyinde karşılaştırılması sonucunda, incelenen alt bölgelerde %89’a kadar ulaşan ve bazı lokal alanlarda daha yüksek oranlara varan yeşil alan azalışları gözlemlenmiştir. Bulgular, afet sonrası yoğun yapılaşma ve kentsel yeniden düzenleme süreçlerinin çevresel sürdürülebilirlik üzerinde anlamlı bir baskı oluşturabileceğini göstermektedir. Önerilen yöntem, afet sonrası çevresel değişimlerin izlenmesi ve sürdürülebilir kentsel planlama stratejilerinin geliştirilmesi için hesaplama açısından verimli ve ölçeklenebilir bir karar destek çerçevesi sunmaktadır.

Anahtar Kelimeler

• Yeşil alan kaybı
• Depremler ve çevresel etkiler
• Uzaktan algılama verileri
• Derin öğrenme
• Anlamsal segmentasyon

Kaynakça

1. Alisjahbana, I., Li, J., & Zhang, Y. (2024). Deepdamagenet: A two-step deep-learning model for multi-disaster building damage segmentation and classification using satellite imagery. arXiv preprint arXiv:2405.04800. https://doi.org/10.48550/arXiv.2405.04800
2. Alpsalaz, S. D., Aslan, E., Özüpak, Y., Alpsalaz, F., Uzel, H., & Bereznychenko, V. (2025). Hybrid deep learning with attention fusion for enhanced colon cancer detection. Scientific Reports. https://doi.org/10.1038/s41598-025-29447-8
3. Alswilem, L., & Pacal, N. (2025). Artificial intelligence in mammography: A study of diagnostic accuracy and efficiency. Computational Systems and Artificial Intelligence, 1(1), 26-31. https://doi.org/10.69882/adba.csai.2025075
4. Aslan, E., & Özüpak, Y. (2025). Detection of road extraction from satellite images with deep learning method. Cluster Computing, 28(1), 72. https://doi.org/10.1007/s10586-024-04880-y
5. Çakmak, M. A., Aslan, E., Demirhan, Y., Esen, F., & Yılmaz, F. K. (2026). A Genetic Algorithm-Based Traffic Light Optimization Model for Efficient Home Healthcare Service Delivery in Türkiye. Artificial Intelligence in Applied Sciences, 2(1), 1-7. https://doi.org/10.69882/adba.ai.2026011
6. Cheng, G., Huang, Y., Li, X., Lyu, S., Xu, Z., Zhao, H., ... & Xiang, S. (2024). Change detection methods for remote sensing in the last decade: A comprehensive review. Remote Sensing, 16(13), 2355. https://doi.org/10.3390/rs16132355
7. Forest Aerial Images for Segmentation dataset (2020). https://www.kaggle.com/datasets/quadeer15sh/augmented-forest-segmentation Date:15.02.2025 Ghafory-Ashtiany, M., & Hosseini, M. (2008). Post-Bam earthquake: recovery and reconstruction. Natural Hazards, 44, 229-241. https://doi.org/10.1007/s11069-007-9108-3
8. Honjo, T., & Takakura, T. (1990). Simulation of thermal effects of urban green areas on their surrounding areas. Energy and buildings, 15(3-4), 443-446. https://doi.org/10.1016/0378-7788(90)90019-F

9. Huang, T. S., Schreiber, W. F., & Tretiak, O. J. (2005). Image processing. Proceedings of the IEEE, 59(11), 1586-1609. https://doi.org/10.1109/PROC.1971.8491
10. Hussain, M., Chen, D., Cheng, A., Wei, H., & Stanley, D. (2013). Change detection from remotely sensed images: From pixel-based to object-based approaches. ISPRS Journal of photogrammetry and remote sensing, 80, 91-106. https://doi.org/10.1016/j.isprsjprs.2013.03.006
11. İnce, S., Kunduracioglu, I., Bayram, B., & Pacal, I. (2025). U-net-based models for precise brain stroke segmentation. Chaos Theory and Applications, 7(1), 50-60. https://doi.org/10.51537/chaos.1605529
12. Kastens, J. H., Kastens, T. L., Kastens, D. L., Price, K. P., Martinko, E. A., & Lee, R. Y. (2005). Image masking for crop yield forecasting using AVHRR NDVI time series imagery. Remote Sensing of Environment, 99(3), 341-356. https://doi.org/10.1016/j.rse.2005.09.010
13. Kunduracioglu, I. (2024). Cnn models approaches for robust classification of apple diseases. Computer and Decision Making: An International Journal, 1, 235-251. https://doi.org/10.59543/comdem.v1i.10957
14. Kunduracioglu, I., & Pacal, I. (2024). Advancements in deep learning for accurate classification of grape leaves and diagnosis of grape diseases. Journal of Plant Diseases and Protection, 131(3), 1061-1080. https://doi.org/10.1007/s41348-024-00896-z
15. Li, M., Zhang, T., Tu, Y., Ren, Z., & Xu, B. (2022). Monitoring Post-Flood Recovery of Croplands Using the Integrated Sentinel-1/2 Imagery in the Yangtze-Huai River Basin. Remote Sensing, 14(3), 690. https://doi.org/10.3390/rs14030690
16. Li, Y., Li, Y., Zhu, X., Fang, H., & Ye, L. (2024.a). A method for extracting buildings from remote sensing images based on 3DJA-UNet3+. Scientific Reports, 14(1), 19067. https://doi.org/10.1038/s41598-024-70019-z
17. Li, Y., Zou, S., Zhao, T., & Su, X. (2024.b). MDFA-Net: Multi-Scale Differential Feature Self-Attention Network for Building Change Detection in Remote Sensing Images. Remote Sensing, 16(18). https://doi.org/10.3390/rs16183466
18. Lissner, I., Preiss, J., Urban, P., Lichtenauer, M. S., & Zolliker, P. (2012). Image-difference prediction: From grayscale to color. IEEE Transactions on Image Processing, 22(2), 435-446. https://doi.org/10.1109/TIP.2012.2216279
19. Liu, P., Zhuang, W., Kou, W., Wang, L., Wang, Q., & Deng, Z. (2025). Fire Severity Outperforms Remote Sensing Indices in Exploring Post-Fire Vegetation Recovery Dynamics in Complex Plateau Mountainous Regions. Forests, 16(2), 263. https://doi.org/10.3390/f16020263
20. Özüpak, Y., Aslan, E., & Zaitsev, I. (2026). Explainable LSTM-AdamW based fault diagnosis of aircraft rotating components using airborne acoustic signals under dynamic operating conditions. Scientific Reports. https://doi.org/10.1038/s41598-026-41889-2
21. Paçal, İ., & Kunduracıoğlu, İ. (2024). Data-efficient vision transformer models for robust classification of sugarcane. Journal of Soft Computing and Decision Analytics, 2(1), 258-271. https://doi.org/10.31181/jscda21202446
22. Pacal, I., & Kunduracioglu, I. (2025). ASMM-Net: A Hybrid CNN with Adaptive Scale Modulation for Robust Ischemic Stroke Lesion Segmentation. Iranian Journal of Science and Technology, Transactions of Electrical Engineering, 1-17. https://doi.org/10.1007/s40998-025-00940-0
23. Pacal, I., Algarni, A., & Kunduracioglu, I. (2026). Adaptive and efficient deep learning model for automated ischemic stroke lesion segmentation. Biomedical Signal Processing and Control, 118, 109658. https://doi.org/10.1016/j.bspc.2026.109658
24. Potere, D. (2008). Horizontal positional accuracy of Google Earth's high-resolution imagery archive. Sensors, 8(12), 7973-7981. https://doi.org/10.3390/s8127973
25. Reilly, T. (2005). An ergonomics model of the soccer training process. Journal of sports sciences, 23(6), 561-572. https://doi.org/10.1080/02640410400021245
26. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. Medical Image Computing and Computer-Assisted Intervention – MICCAI, 234–241. https://doi.org/10.1007/978-3-319-24574-4_28
27. Soleimani-Babakamali, M. H., Askari, M., Heravi, M. A., Sisman, R., Attarchian, N., Askan, A., ... & Taciroglu, E. (2025). Deep Ensemble Learning for Rapid Large-Scale Postearthquake Damage Assessment: Application to Satellite Images from the 2023 Türkiye Earthquakes. ASCE OPEN: Multidisciplinary Journal of Civil Engineering, 3(1), 04025003. https://doi.org/10.1061/AOMJAH.AOENG-0043
28. Sykes, L. R. (1967). Mechanism of earthquakes and nature of faulting on the mid‐oceanic ridges. Journal of Geophysical Research, 72(8), 2131-2153. https://doi.org/10.1029/JZ072i008p02131
29. Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016). Rethinking the Inception Architecture for Computer Vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2818–2826.
30. Tavus, B. E. S. T. E., Kocaman, S., Nefeslioglu, H. A., & Gokceoglu, C. (2020). A fusion approach for flood mapping using Sentinel-1 and Sentinel-2 datasets. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 43, 641-648. https://doi.org/10.5194/isprs-archives-XLIII-B3-2020-641-2020
31. Tong, G., Li, Y., Gao, H., Chen, H., Wang, H., & Yang, X. (2020). MA-CRNN: a multi-scale attention CRNN for Chinese text line recognition in natural scenes. International Journal on Document Analysis and Recognition (IJDAR), 23(2), 103-114. https://doi.org/10.1007/s10032-019-00348-7 Wang, L., Sousa, W. P., & Gong, P. (2004). Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery. International journal of remote sensing, 25(24), 5655-5668. https://doi.org/10.1080/014311602331291215
32. Young, S., Balluz, L., & Malilay, J. (2004). Natural and technologic hazardous material releases during and after natural disasters: a review. Science of the total environment, 322(1-3), 3-20. https://doi.org/10.1016/S0048-9697(03)00446-7
33. Yu, F., & Koltun, V. (2015). Multi-scale context aggregation by dilated convolutions. arXiv preprint arXiv:1511.07122. https://doi.org/10.48550/arXiv.1511.07122
34. Yu, L., & Gong, P. (2012). Google Earth as a virtual globe tool for Earth science applications at the global scale: progress and perspectives. International Journal of Remote Sensing, 33(12), 3966-3986. https://doi.org/10.1080/01431161.2011.636081
35. Zhang, C., Sargent, I., Pan, X., Li, H., Gardiner, A., Hare, J., & Atkinson, P. M. (2019). Joint Deep Learning for land cover and land use classification. Remote sensing of environment, 221, 173-187. https://doi.org/10.1016/j.rse.2018.11.014
36. Zhang, X., Li, L., Di, D., Wang, J., Chen, G., Jing, W., & Emam, M. (2022). SERNet: Squeeze and excitation residual network for semantic segmentation of high-resolution remote sensing images. Remote Sensing, 14(19), 4770. https://doi.org/10.3390/rs14194770 Zhang, Y. L., Kim, Y. C., & Cha, G. W. (2023). Assessment of deep learning-based image analysis for disaster waste identification. Journal of Cleaner Production, 428, 139351. https://doi.org/10.1016/j.jclepro.2023.139351
37. Zhang, Y., Tang, X., Liu, D., Taymaz, T., Eken, T., Guo, R., ... & Sun, H. (2023). Geometric controls on cascading rupture of the 2023 Kahramanmaraş earthquake doublet. Nature Geoscience, 16(11), 1054-1060. https://doi.org/10.1038/s41561-023-01283-3
38. Zhao, B., Sui, H., Liu, J., Shi, W., Wang, W., Xu, C., & Wang, J. (2024). Flood inundation monitoring using multi-source satellite imagery: A knowledge transfer strategy for heterogeneous image change detection. Remote Sensing of Environment, 314, 114373. https://doi.org/10.1016/j.rse.2024.114373.