Assessment of the macroseismic/strong-ground motion distribution of the 2021 Crete (Arkalochori) earthquake sequence using a finite fault stochastic simulation approach

<p>At 27/09/2021, 06:17 (UTC) a strong ground motion with moment magnitude <strong>M</strong>6.0 occurred on the island of Crete, approximately 25km SE of the city of Heraklion, near Arkalochori. The highest macroseismic intensity value was observed in the area of the central part of the peripheral unit of Heraklion (i.e., in the area of the Municipality of Minoa Pediada) and had a value of I<sub>MM</sub> = VII. The earthquake was also felt in the islands of the southern and eastern Aegean up to areas of Attica. We collected macroseismic data from EMSC considering a significant number of macroseismic testimonies and available strong motion information. The main goal was to perform a combined interpretation between observed and synthetic macroseismic data. In order to predict the expected ground motion measurements, for example peak ground acceleration (PGA) and peak ground velocity (PGV), as a function of distance and magnitude we used the stochastic simulation approach. These simulations are performed with the EXSIM code (Motazedian and Atkinson, 2005), as described by Boore (2009) taking into account finite-fault effects in ground-motion modeling. Good knowledge of the detailed rupture process is essential for realistic simulations of strong ground motion. Earthquake relocations for this aftershock sequence suffer from poor knowledge of the local velocity structure, especially for the shallow part of the crust. This was an important factor in the case of this earthquake, as the permanent network is rather sparse in this area. We employed a Monte Carlo parametric search of the velocity model space, realized through an adapted neighborhood algorithm, as included in the Geopsy software, together with a conventional location code. In this approach, the regional 1D velocity model, together with appropriate station corrections, is simultaneously estimated (non-linear optimization) with the relocation of the complete seismic sequence. Finally, a good agreement of the spatial distribution of the initial and modeled simulated macroseismic intensities is observed, showing that can reliably reconstruct the main features of the damage distribution approach for this earthquake.</p><p> </p>