Impact of fault modeling assumptions on regional seismic risk assessment: A case study of the 1963 Skopje earthquake, North Macedonia

North Macedonia lies near the boundary between the Eurasian and African tectonic plates. The movement of these tectonic plates leads to frequent moderate to strong earthquakes, making the Balkan region one of the most seismically active areas in the Mediterranean. In North Macedonia, the Skopje earthquake of July 26, 1963, was one of the most destructive events, causing significant damage and losses to the region's building inventory. It struck Skopje, the most densely populated area of the country.

The Skopje earthquake has been extensively studied. However, it has been observed that input parameters (i.e., depth, magnitude, fault kinematics) used in fault modeling vary between different sources, highlighting the lack of consistency in the models used to characterize seismic hazard on the region. These uncertainties are related to the limited availability of seismological information, a problem that generally affects many historical earthquakes. Since fault characterization has a significant impact on surface ground motion and, consequently, on the seismic risk assessment, particular attention must be given to the input parameters used in scenario modeling.

This study aims to assess the impact of different modeling approaches regarding the depth, magnitude, and kinematics of the 1963 Skopje earthquake on seismic risk evaluation of the Skopje region. Particular attention is given to the selection of Ground Motion Models (GMMs), as some may not be sensitive to certain parameters under consideration. The exposure model is developed using aggregated data from North Macedonia's latest census conducted in 2021. Vulnerability models for the building typologies identified in the census are derived from the GEM vulnerability database and implemented through a GIS scheme. The uncertainties associated with the exposure and vulnerability models are briefly addressed.

The hazard and risk analyses are carried out using the state-of-the-art software, the OpenQuake Engine, and the risk analysis results are ultimately presented in terms of damages to the building inventory, as well as direct and indirect losses.