The 6 February 2023 Türkiye Earthquakes: Insights for the European Seismic Hazard and Risk Models

Abstract

The earthquakes that struck eastern Türkiye and Syria on 6 February 2023, first with a Mw 7.8 shock then followed only hours later by a second Mw 7.6 event, will have profound and long-lasting consequences for those living in this highly seismically active region. From the perspective of the European earthquake science and engineering communities, however, these events also force us to evaluate our models of seismic hazard and risk for the region, specifically the 2020 European Seismic Hazard Model (ESHM20, Danciu et al.. 2021) and European Seismic Risk Model (ESRM20, Crowley et al.. 2021), to identify potential shortcomings and focus on areas where improvement is needed. A single event such as this can neither validate nor invalidate probabilistic models, but as data emerge, we can compare these with components of our models and verify the extent to which the events themselves and their consequences are described.We first verified that the ruptures associated to the two main earthquakes are present within the inventory of ruptures and associated probabilities within the source model (the earthquake rupture forecasts or ERFs) for the East Anatolian Fault (first event) and Sürgü-Cardak Fault (second event). These earthquakes are larger than those in the historical earthquake catalogue, but ruptures close in magnitude and dimension to those observed were present in the ESHM20 ERFs. Both magnitudes were between 0.2 – 0.4 Mw units lower than those defined for their respective faults on the different logic tree branches.Preliminary ground motion observations allowed us to compare the observed shaking to that predicted by the ESHM20 ground motion model (GMM) and others in the literature. These were found to be consistent in their prediction of the expected shaking and its attenuation. The 6th February earthquakes do show that future models must address issues of time-dependence between earthquakes and allow for short-term clustering of large events on nearby ruptures. Recorded near-fault ground motions also suggest strong pulse-like behaviour, indicating the need for such phenomena to be better captured in the GMMs.A complete assessment of the actual damage and consequences is not yet available from which we could compare the seismic risk model. We have run scenario risk calculations using the ESRSM20 site, exposure and vulnerability models for the two main earthquakes, along with other scenario ruptures on neighbouring faults. Expected fatalities were lower than those reported at the time of writing; however, many factors contribute to this. Further analysis is needed to understand the difference, but critical areas for future improvement to the risk models should include state-dependent fragility, modelling of further epistemic uncertainty in exposure and vulnerability, and inclusion of spatial- and temporal correlations in ground motions across a region. Future efforts by the seismic hazard and risk modelling community to address these issues considering the February 2023 earthquakes may have a lasting impact on risk mitigation, both in Türkiye and across Europe.