Probabilistic seismic hazard analysis for Croatia: hazard mapping and Eurocode 8 implications

This study presents a probabilistic seismic hazard analysis (PSHA) for Croatia and its 3 major tectonic zones: Adriatic Foreland, Dinarides, and Pannonian Basin, using the OpenQuake Engine in classical mode, with emphasis on the construction, justification and validation of a Croatia-specific seismic source model and its implications for seismic design practice. The seismic source model comprises 32 area zones grouped into nine seismotectonic super-zones with Gutenberg–Richter parameters for each zone estimated using the Weichert maximum likelihood method and validated against independent seismicity observations. This source model, rather than the computational workflow used to process it, is the study’s primary scientific contribution; the OpenQuake implementation includes a consistently weighted ground-motion logic tree based on a combined DIC + EDR ranking, including a Kowsari region-specific model implemented in a custom Python framework. To the authors’ knowledge, this is the first PSHA for Croatia to systematically combine a regionally calibrated ground-motion model with a formally ranked logic-tree framework and to provide a full comparison against Eurocode 8 (EC8) design spectra. Hazard was computed for PGA and spectral accelerations between 0.1 and 2.0 s, on a 0.1° grid, for seven return periods ranging from 95 to 2475 years. The resulting spatial pattern reflects the underlying source model: hazard peaks along the Adriatic coast, where zone ADR3 - identified as the highest-hazard zone - makes the largest contribution. A systematic comparison with Eurocode 8 spectra from the Croatian National Annex highlights a previously undocumented, period-dependent discrepancy: PSHA spectra exceed code values at short periods but fall below them at longer periods. This finding provides new insight into the applicability of current design spectra and directly informs the ongoing revision of the national seismic hazard and design framework.