The Vrancea seismic zone (VSZ), located in Romania, at the sharp bend of the southeast Carpathians, is an anomalous intraplate seismic nest releasing the largest strain in continental Europe. The last powerful earthquake in Vrancea occurred in 1977 causing significant damage to the densely populated cities in southeast Romania. The seismic infrastructure was underdeveloped at that time and the earthquake was not well recorded locally. Using ambient seismic noise recorded at modern broadband seismic stations around Vrancea in the year 2020, and the approximate moment tensor solutions of significant earthquakes Mw>6 since 1977, we reconstruct ground-motion waveforms recorded by modern seismometers decades after their occurrence. The virtual earthquake approach uses empirical Green’s functions from ambient noise cross correlations between pairs of seismic stations and adds the signatures of a model earthquake: double couple mechanism, buried source, and a realistic earth model in the epicentral area. Ambient noise data capture the signatures of complex subsurface structures and the 3D path effects, providing a more physically accurate representation of the seismic wavefield than purely synthetic data. Our new results demonstrate the viability of this innovative method and provide a unique opportunity for more accurate seismic hazard analysis in intraplate seismic zones with insufficient instrumental data.

Evaluating the performance of probabilistic seismic hazard models against recorded data and their potential to forecast future earthquake’s ground shaking is an emerging research topic. In this study, we evaluate and test the results of the recently released European Seismic Hazard Model (ESHM20; Danciu et al., 2021) against observations for several cities in Romania. The dataset consists of ground shaking recordings and macroseismic observations, which extend the observational time-period to few hundred years. The full distribution of the hazard curves, depicting the epistemic uncertainties of the hazard at the given location was considered and the testing was done for peak ground acceleration (PGA) values i.e., 0.1 g and 0.2 g. The results show close agreement between the ESHM20 and the observations for the cities located near the Vrancea intermediate-depth source (VRI) for both selected PGA levels. ESHM20 appears to overestimate the VRI recorded ground motions along the Carpathian Mountain Range and underestimate those at the far-field locations outside the Carpathians. Some of these differences might be attributed to the uncertainties in data conversion, local site effects, or differences in the attenuation patterns of the ground motion models. Our analysis suggests that the observed exceedance rates for the selected PGA levels are consistent with ESHM20 estimates, but these results must be interpreted with caution given the limited time and spatial coverage of the observations.