Assessment of building damage from the 2020 Sivrice earthquake using a satellite based rapid seismic screening method

Apart from inducing extensive damage rapidly across vast areas, earthquakes also have enduring consequences as they elevate stress levels in the neighboring regions where seismic activity persists. Although later destructive earthquakes, such as the February 6, 2023 Kahramanmaraş events, further highlighted the seismic vulnerability of the region, this study specifically focuses on assessing the impacts of the January 24, 2020 Sivrice Earthquake (Mw 6.7) on residential areas. After expert assessments of post-earthquake building damage, geolocated and damage-classified buildings were further analyzed through optical and Synthetic Aperture Radar (SAR) satellite imagery. Land classification was performed using optical images and a comparison was made between land class changes and building coordinates to detect collapsed buildings. Collapsed buildings were identified with 89.1% accuracy by analyzing land-use class changes from settlement to bare land between pre- and post-earthquake SVM-based classifications of optical images. Advanced Interferometric Synthetic Aperture Radar (InSAR) techniques, specifically Persistent Scatterer (PS) and Small Baseline Subset (SBAS) methods, were employed using SAR satellite imagery to calculate buildings' annual surface velocities for pre-earthquake dates. Correlation analyses were performed to assess the relationship between building damage classes and the key parameters of epicentral distance of the building, fault distance, the Earthquake Design Class (EDC), geological formation (i.e., Vs30), building type (RC or masonry), number of storeys, surface velocity, and Coulomb stress in order to predict building damage from possible earthquakes in advance by using pre-earthquake data and remote sensing (RS) methods. The damage estimation map was generated by integrating RS and geographic information system (GIS) methodologies, which included results of the correlation analysis that highlighted the dominance of the fault distance of the building, number of storeys, EDC, and surface velocity influencing damage severity. Thus, for seismic risk prioritization of buildings, RS and GIS-based first-stage assessment methods can be deployed as a rapid alternative to Rapid Visual Screening (RVS) methods since the need for one-by-one data collection from buildings by site survey is no longer necessary.