Automated detection of damaged buildings in post-disaster scenarios: a case study of Kahramanmaras (Turkiye) earthquakes on February 6, 2023

ŞERİFOĞLU YILMAZ Ç., YILMAZ V., Tansey K., Aljehani N. S. O.

NATURAL HAZARDS, no.3, pp.1247-1271, 2023 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Publication Date: 2023
  • Doi Number: 10.1007/s11069-023-06154
  • Journal Name: NATURAL HAZARDS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, IBZ Online, PASCAL, Aerospace Database, Agricultural & Environmental Science Database, Aquatic Science & Fisheries Abstracts (ASFA), CAB Abstracts, Communication Abstracts, Environment Index, Geobase, INSPEC, Metadex, PAIS International, Pollution Abstracts, Veterinary Science Database, DIALNET, Civil Engineering Abstracts
  • Page Numbers: pp.1247-1271
  • Karadeniz Technical University Affiliated: Yes


This study develops a novel approach for identifying buildings that were damaged in the aftermath of the Kahramanmaras earthquakes on February 6, 2023, which were among the most devastating in the history of Turkiye. The approach involves using two pre-event and one post-event Sentinel-1 and Sentinel-2 images to detect changes in the varying-sized and shaped buildings following the earthquakes. The approach is based on the hypothesis that the radiometric characteristics of building pixels should change after an earthquake, and these changes can be detected by analysing the spectral distance between the building pixel vectors before and after the earthquake. The proposed approach examines the changes in building pixel vectors on pre-event and post-event Sentinel-2 MultiSpectral Instrument images. It also incorporates the backscattering features of Sentinel-1 Synthetic Aperture Radar images, as well as the variance image, a feature that is derived from a Grey-Level Co-occurrence Matrix, and the Normalized Difference Built-up Index image, which were derived from the optical data. The approach was tested on three sites, two of which were in Kahramanmaras and the third in Hatay city. The results showed that the proposed method was able to accurately identify damaged and undamaged buildings with an overall accuracy of 75%, 84.4%, and 73.8% in test sites 1, 2, and 3, respectively. These findings demonstrate the potential of the proposed approach to effectively identify damaged buildings in post-disaster situations.