A Multi-Sensor Semi-Supervised and Unsupervised Framework for Post-Disaster Flood and Building Damage Assessment: The Case of the Derna Dam Collapse

Abstract

The catastrophic collapse of the Derna Dam created an urgent need for rapid and reliable mapping of flood extent and building damage to support disaster response and recovery. This study presents a multi-sensor, multi-method change detection framework integrating open-access Sentinel-1 Synthetic Aperture Radar (SAR), Sentinel-2 optical imagery, and Very High-Resolution (VHR) Maxar imagery to detect post-event changes and assess building-level damage. Two methodological approaches were evaluated: semi-supervised and unsupervised. The semi-supervised pipeline utilizes a Multilayer Perceptron (MLP)-based post-classification to generate pseudo-labels for training two U-Net variants: single-encoder early fusion and Siamese double-encoder mid-fusion architectures. The unsupervised methods include Principal Component Analysis (PCA) with Change Vector Analysis (CVA), Kernel Canonical Correlation Analysis (KCCA) implemented efficiently with Random Fourier Features (RFF), and a novel PCA with KCCA hybrid that combines dimensionality reduction with nonlinear correlation analysis. Rapid assessment using Sentinel-2 and OpenStreetMap (OSM) data identified changed areas through Change/No-Change and Normalized Difference Moisture Index (NDMI) masks. A detailed analysis employed VHR Maxar imagery and a pretrained Mask Region-based Convolutional Neural Network (Mask R-CNN) within ArcGIS Pro to extract pre- and post-event building footprints and evaluate structural impacts. Validation against manually labeled VHR samples (131 polygons; 3,371 pixels) confirmed the robustness of the framework. Stacked Sentinel-1 + Sentinel-2 fusion improved pseudo-label quality and segmentation accuracy (semi-supervised Overall Accuracy (OA) ≈95.1%), while the PCA＋KCCA hybrid achieved the best unsupervised performance (OA ≈89.4%). For the assessment of building damage, two methods were utilized: Quick Assessment (QA) and Deliberate Assessment (DA). A total of 2,193 flooded buildings were identified using the Deliberate Assessment, while 2,694 flooded buildings were identified through the Quick Assessment. These findings were validated using data from the United Nations Satellite Center (UNOSAT). The study contributes (i) a reproducible, end-to-end workflow for rapid post-disaster mapping, (ii) an efficient RFF-based KCCA implementation, and (iii) a novel PCA＋KCCA hybrid that optimizes the balance between accuracy, robustness, and computational efficiency for operational-scale change detection.

Keywords

• Sentinel-1
• Sentinel-2
• U-NET
• PCA_KCCA
• Derna flood

References

1. (USGS), U. S. G. S. (2025). Normalized Difference Moisture Index (NDMI). Retrieved from https://www.usgs.gov/landsat-missions/normalized-difference-moisture-index
2. Adugna, T., Xu, W., & Fan, J. (2022). Comparison of Random Forest and Support Vector Machine Classifiers for Regional Land Cover Mapping Using Coarse Resolution FY-3C Images. Remote Sensing, 14, 574. doi:10.3390/rs14030574
3. Aimaiti, Y., Sanon, C., Koch, M., Baise, L. G., & Moaveni, B. (2022). War Related Building Damage Assessment in Kyiv, Ukraine, Using Sentinel-1 Radar and Sentinel-2 Optical Images. Remote Sensing, 14(24). doi:10.3390/rs14246239
4. Akgül, M. A. (2018). Sentetik Açıklıklı Radar verilerinin Taşkın Çalışmalarında Kullanılması: Berdan Ovası Taşkını. Geomatik, 3(2), 154-162. doi:10.29128/geomatik.378123
5. Aldoski, J., Mansor, S., Shafri, H., & Shafri, M. (2013). Change Detection Process and Techniques. ResearchGate.
6. Alzate Atehortúa, B. E., & Valero Medina, J. A. (2019). Comparison of maximum likelihood, support vector machines, and random forest techniques in satellite images classification. Tecnura, 23(59), 13-26. doi:10.14483/22487638.14826
7. Anirudh, R., Jungkyo, J., Vítor, S., Giuseppe, M., & Sang‐Ho, Y. (2023). Earthquake building damage detection based on synthetic-aperture-radar imagery and machine learning. Natural Hazards and Earth System Sciences. doi:10.5194/nhess-23-789-2023
8. Armon, M., Shmilovitz, Y., & Dente, E. (2025). Anatomy of a foreseeable disaster: Lessons from the 2023 dam-breaching flood in Derna, Libya. Science Advances, 11(13), eadu2865. doi:doi:10.1126/sciadv.adu2865

9. Bakırman, T., Aslantaş, N., Selbesoğlu, M. O., & Bayram, B. (2025). The Role of Ensemble Deep Learning for Building Extraction from VHR Imagery. International Journal of Engineering and Geosciences, 10(3), 352-363. doi:10.26833/ijeg.1587798
10. Brian McWilliams, u., Switzerland, David Balduzzi, urich, Switzerland, Joachim M. Buhmann, urich, Switzerland. (2013). Correlated random features for fast semi-supervised learning.
11. Carvalho Júnior, O. A., Guimarães, R. F., Gillespie, A. R., Silva, N. C., & Gomes, R. A. T. (2011). A New Approach to Change Vector Analysis Using Distance and Similarity Measures. Remote Sensing, 3(11), 2473-2493. doi:10.3390/rs3112473
12. Chen Wu, M., IEEE, Hongruixuan Chen, Bo Du, Senior Member, IEEE, and Liangpei Zhang, Fellow, IEEE. (2019). Unsupervised Change Detection in Multi-temporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network. Paper presented at the arXiv.
13. 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). doi:10.3390/rs16132355
14. Cho, S., Ma, J., & Yakimenko, O. A. (2023). Aerial multi-spectral AI-based detection system for unexploded ordnance. Defence Technology, 27, 24-37. doi:https://doi.org/10.1016/j.dt.2022.12.002
15. Dapke, P., Quadri, S. A., Nagare, S. M., Bandal, S. B., & Baheti, M. R. (2025). A Comparative Analysis Machine Learning Techniques for LULC Classification Using Landsat-8 Satellite Imagery. International Journal of Engineering and Geosciences, 10(1), 84-92. doi:10.26833/ijeg.1503104
16. David Lopez-Paz , S. S., Alexander J. Smola, Zoubin Ghahramani, Bernhard Scholkopf (2014). Randomized Nonlinear Component Analysis.
17. Demir, V., Yılmaz, C. B., Bodu, H., Yüce, E. S., & Sevimli, M. F. (2023). Türkiye’nin uzun dönem ortalama sıcaklık (°C) değerlerinin üç farklı enterpolasyon yöntemi ile tahmini. Geomatik, 8(1), 9-17. doi:10.29128/geomatik.984310
18. Di Martino, T., Le Saux, B., Guinvarc’h, R., Thirion-Lefevre, L., & Colin, E. (2023). Detection of Forest Fires through Deep Unsupervised Learning Modeling of Sentinel-1 Time Series. ISPRS International Journal of Geo-Information, 12(8), 332. Retrieved from https://www.mdpi.com/2220-9964/12/8/332
19. Duysak, H., & Yiğit, E. (2022). Investigation of the performance of different wavelet-based fusions of SAR and optical images using Sentinel-1 and Sentinel-2 datasets. International Journal of Engineering and Geosciences, 7(1), 81-90. doi:10.26833/ijeg.882589
20. El-bouhalı, A., Amyay, M., & El Ouazanı Ech-chahdı, K. (2025). Detection of irrigated crops using Landsat 8 images: A methodology based on fieldwork and spectral reflectance analysis. International Journal of Engineering and Geosciences, 10(1), 1-13. doi:10.26833/ijeg.1483206
21. Emmanouil, P., Konstantinos, X. S., Melpomeni, Z., & Nicholas, D. (2019). Synergistic Approach of Remote Sensing and GIS Techniques for Flash-Flood Monitoring and Damage Assessment in Thessaly Plain Area, Greece. Water. doi:10.3390/w11030448
22. Erener, A., & Sarp, G. (2017). Barajların Çevresel Etkilerinin Zamansal ve Mekansal Olarak Uzaktan Algılama İle Değerlendirilmesi: Atatürk Barajı Örneği. [Barajların Çevresel Etkilerinin Zamansal Ve Mekansal Olarak Uzaktan Algılama İle Değerlendirilmesi: Atatürk Barajı Örneği]. Geomatik, 2(1), 1-10. doi:10.29128/geomatik.300012
23. Esri. (2025). Deep learning model to extract building footprints in Africa from high-resolution aerial or satellite imagery. . Retrieved from https://www.arcgis.com/home/item.html?id=979cb0cf938946bfb8bb2f41cf9f9795
24. Exchange, T. H. D. (2023). Libya: Building Footprints in Derna Flood Area. Retrieved from https://data.humdata.org/dataset/buildings-footprint-floods-derna-libya
25. Forkuor, G., Dimobe, K., Serme, I., & Tondoh, J. E. (2017). Landsat-8 vs. Sentinel-2: examining the added value of sentinel-2’s red-edge bands to land-use and land-cover mapping in Burkina Faso. GIScience & Remote Sensing, 55(3), 331-354. doi:10.1080/15481603.2017.1370169
26. Gao, B.-c. (1996). NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sensing of Environment, 58(3), 257-266. doi:https://doi.org/10.1016/S0034-4257(96)00067-3
27. HUANG Wei1, HUANG Jinliang1, WANG Lihui1, HU Yanxia1,2, HAN Pengpeng1,2. (2016). Remote sensing image change detection based on change vector analysis of PCA component. REMOTE SENSING FOR LAND & RESOURCES.
28. Hub, S. (2025). Documentation page on NDMI. Retrieved from https://custom-scripts.sentinel-hub.com/custom-scripts/sentinel-2/ndmi/
29. Jiawei, T., Shengquan, Y., Shujuan, H., & Bozhi, X. (2024). Remote Sensing Building Damage Assessment Based on Machine Learning. International Journal of Advanced Network, Monitoring, and Controls. doi:10.2478/ijanmc-2024-0021
30. Nigar, A., Li, Y., Jat Baloch, M. Y., Alrefaei, A. F., & Almutairi, M. H. (2024). Comparison of machine and deep learning algorithms using Google Earth Engine and Python for land classifications. Frontiers in Environmental Science, 12. doi:10.3389/fenvs.2024.1378443
31. Pilario, K. E., Shafiee, M., Cao, Y., Lao, L., & Yang, S.-H. (2019). A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring. Processes, 8(1). doi:10.3390/pr8010024
32. Polykretis, C., Grillakis, M., & Alexakis, D. (2020). Exploring the Impact of Various Spectral Indices on Land Cover Change Detection Using Change Vector Analysis: A Case Study of Crete Island, Greece. Remote Sensing, 12(2). doi:10.3390/rs12020319
33. Recht, A. R. a. B. (2007). Random Features for Large-Scale Kernel Machines. NIPS.
34. ROAS, U.-h. (Cartographer). (2023). Derna damage map. Retrieved from https://unhabitat- roas.hub.arcgis.com/maps/unhabitat- roas::derna- damage- map/about.
35. Shafique, A., Cao, G., Khan, Z., Asad, M., & Aslam, M. (2022). Deep Learning-Based Change Detection in Remote Sensing Images: A Review. Remote Sensing, 14(4). doi:10.3390/rs14040871
36. Sudiana, D., Lestari, A. I., Riyanto, I., Rizkinia, M., Arief, R., Prabuwono, A. S., & Sri Sumantyo, J. T. (2023). A Hybrid Convolutional Neural Network and Random Forest for Burned Area Identification with Optical and Synthetic Aperture Radar (SAR) Data. Remote Sensing, 15(3). doi:10.3390/rs15030728
37. UNOSAT. (2023). Flood impact assessment in Derna City, East Province, Libya. FL20230912LBY. Retrieved from https://unosat.org/products/3671
38. Volpi, M., Camps-Valls, G., & Tuia, D. (2015). Spectral alignment of multi-temporal cross-sensor images with automated kernel canonical correlation analysis. ISPRS Journal of Photogrammetry and Remote Sensing, 107, 50-63. doi:10.1016/j.isprsjprs.2015.02.005
39. Wilson, E. H., & Sader, S. A. (2002). Detection of forest harvest type using multiple dates of Landsat TM imagery. Remote Sensing of Environment, 80(3), 385-396. doi:https://doi.org/10.1016/S0034-4257(01)00318-2
40. Yinsong Wang and Shahin Shahrampour, S. M., IEEE. (2021). ORCCA: Optimal Randomized CanonicalCorrelation Analysis. Paper presented at the arXiv:1910.05384v3.
41. You, Y., Cao, J., & Zhou, W. (2020). A Survey of Change Detection Methods Based on Remote Sensing Images for Multi-Source and Multi-Objective Scenarios. Remote Sensing, 12(15). doi:10.3390/rs12152460
42. Youssef, A. E., Elsharkawy, A. S., Mosa, O. M., & Tarek, M. (2025). Land Use and Land Cover Change Detection in the Gaza Strip Using Sentinel-2: Integrating Enhanced Lee Filter with Artificial Neural Networks. Journal of Engineering Science and Military Technologies, -. doi:10.21608/ejmtc.2025.357322.1303
43. Zhang, M., Liu, Z., Feng, J., Liu, L., & Jiao, L. (2023). Remote Sensing Image Change Detection Based on Deep Multi-Scale Multi-Attention Siamese Transformer Network. Remote Sensing, 15(3). doi:10.3390/rs15030842
44. Zhu, X. X., Tuia, D., Mou, L., Xia, G. S., Zhang, L., Xu, F., & Fraundorfer, F. . (2017). Deep Learning in Remote Sensing: A Review. IEEE Geoscience and Remote Sensing Magazine.