The 6.4 magnitude Petrinja earthquake, which struck central Croatia on December 29, 2020, was the most powerful recorded in the country. To assess its impact, we employed Interferometric Synthetic Aperture Radar (InSAR) and Polarimetric Synthetic Aperture Radar (PolSAR) to examine surface deformation and building damage. We used descending Sentinel-1 Single-Look Complex (SLC) images from the earthquake and compared them with data from June to account for the influence of winter conditions. InSAR effectively detected deformation by combining pre- and post-earthquake datasets, followed by an analysis of the affected area based on the unwrapped image profile. PolSAR analysis, using the unsupervised Wishart polarimetric decomposition method, classified data into three categories: buildings, vegetation, and water bodies. Multiple iterations (3, 25, 50, 75, and 100) were tested to identify optimal results. While InSAR accurately detected deformation, PolSAR showed lower accuracy, particularly when comparing pre- and post-earthquake data. After repeated testing, the optimal PolSAR iterations were 100 for pre-earthquake data and 25 for post-earthquake and June data. The analysis revealed a reduction in building area from 2,738.38 hectares before the earthquake to 2,063.38 hectares afterward. In conclusion, the earthquake reduced the building area, and PolSAR performance was less reliable under winter conditions.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Integrating InSAR Profiling and Unsupervised Wishart PolSAR for Assessing Earthquake-Induced Building Damage

  • Contardo Ferrini Abedidan Andreavalent Santosa,
  • Naufal Setiawan,
  • Thema Arrisaldi,
  • Lysa Dora Ayu Nugraini,
  • Syachrul Arief

摘要

The 6.4 magnitude Petrinja earthquake, which struck central Croatia on December 29, 2020, was the most powerful recorded in the country. To assess its impact, we employed Interferometric Synthetic Aperture Radar (InSAR) and Polarimetric Synthetic Aperture Radar (PolSAR) to examine surface deformation and building damage. We used descending Sentinel-1 Single-Look Complex (SLC) images from the earthquake and compared them with data from June to account for the influence of winter conditions. InSAR effectively detected deformation by combining pre- and post-earthquake datasets, followed by an analysis of the affected area based on the unwrapped image profile. PolSAR analysis, using the unsupervised Wishart polarimetric decomposition method, classified data into three categories: buildings, vegetation, and water bodies. Multiple iterations (3, 25, 50, 75, and 100) were tested to identify optimal results. While InSAR accurately detected deformation, PolSAR showed lower accuracy, particularly when comparing pre- and post-earthquake data. After repeated testing, the optimal PolSAR iterations were 100 for pre-earthquake data and 25 for post-earthquake and June data. The analysis revealed a reduction in building area from 2,738.38 hectares before the earthquake to 2,063.38 hectares afterward. In conclusion, the earthquake reduced the building area, and PolSAR performance was less reliable under winter conditions.