Search

The 2023 February 6 Mw 7.8 earthquake was the first one of a doublet which shook Türkiye and Syria causing, as per the estimates at the time of writing of this abstract, more than 45,000 casualties.The current standard operating procedures of the NEAMTWS (Tsunami Warning System in the North-Eastern Atlantic, the Mediterranean and connected seas, coordinated by UNESCO/IOC) for the initial tsunami warning message following an earthquake are based on a Decision Matrix (DM), whose input parameters are hypocentre and magnitude of the earthquake. Since the epicentre of this earthquake was located at a depth between 15-35 km at almost 100 km from the coast, both KOERI (Türkiye) and INGV (Italy) Tsunami Service Providers (TSPs) of the NEAMTWS issued a Tsunami Watch message (i.e., runup expected to exceed 1 m) for the whole Mediterranean Sea. NOA (Greece) did not issue any alert, because its initial location was more than 100 km from the coast.In response to the tsunami warning, trains were stopped in different locations in Southern Italy for several hours, and evacuation of some coastal areas was enforced. However, only a relatively small tsunami was recorded by Turkish close-by tide-gauges in the Eastern Mediterranean, with a maximum recorded amplitude of less than 50 cm. Based on these measurements and on others showing little to no tsunami at increasing distances, the alert was then ended after 5 and 9 hours by INGV and KOERI, respectively, based on the available tide-gauge recordings and interaction with Civil Protection Officers.This event has highlighted that NEAMTWS is an asset for the coastal communities. It can provide rapid alerts, which can save lives if the last-mile of the procedures is in place and the communities are “Tsunami Ready”, that is aware and prepared to respond with evacuations and other appropriate countermeasures. On the other hand, while it is reasonable – and dutiful based on current standard operation procedures – to issue a basin-wide, or at least a local alert, for an inland earthquake of unknown mechanism and of such a large magnitude, it is perhaps possible to improve the DM, which is totally heuristic and characterized by hard-thresholds, with consideration of numerical tsunami simulations and quantitative uncertainty treatment with more continuous variations. Moreover, there is no procedure currently in place to differentiate among locations where the expected time of arrival differs by many hours across the Mediterranean basin, nor a sufficient instrumental coverage that could make cancellation/ending faster due to a more solid observational basis.We will discuss some of the scientific and operational aspects with the aim of identifying which lessons can be learned to improve the NEAMTWS efficiency. We will also compare the DM-based alerts with those that would be produced with the recently introduced Probabilistic Tsunami Forecasting (PTF, Selva et al., 2021, Nature Communications), presently in pre-operational testing at INGV.

We present a source solution for the tsunami generated by the Mw 6.6 earthquake that occurred on 2 May 2020, about 80 km offshore south of Crete, in the Cretan Passage, on the shallow portion of the Hellenic Arc subduction zone (HASZ). The tide gauges recorded this local tsunami on the southern coast of Crete and Kasos island. We used Crete tsunami observations to constrain the geometry and orientation of the causative fault, the rupture mechanism, and the slip amount. We first modelled an ensemble of synthetic tsunami waveforms at the tide gauge locations, produced for a range of earthquake parameter values as constrained by some of the available moment tensor solutions. We allow for both a splay and a back-thrust fault, corresponding to the two nodal planes of the moment tensor solution. We then measured the misfit between the synthetic and the Ierapetra observed marigram for each source parameter set. Our results identify the shallow, steeply dipping back-thrust fault as the one producing the lowest misfit to the tsunami data. However, a rupture on a lower angle fault, possibly a splay fault, with a sinistral component due to the oblique convergence on this segment of the HASZ, cannot be completely ruled out. This earthquake reminds us that the uncertainty regarding potential earthquake mechanisms at a specific location remains quite significant. In this case, for example, it is not possible to anticipate if the next event will be one occurring on the subduction interface, on a splay fault, or on a back-thrust, which seems the most likely for the event under investigation. This circumstance bears important consequences because back-thrust and splay faults might enhance the tsunamigenic potential with respect to the subduction interface due to their steeper dip. Then, these results are relevant for tsunami forecasting in the framework of both the long-term hazard assessment and the early warning systems.

This study presents the first tsunami scenario database in Marmara Sea, Turkey referring to 30 different earthquake scenarios obtained with the combinations of 32 possible fault segments. The fault mechanisms in Marmara Sea have been studied in detail within FP-7 MARSite project, which were derived from various databases and literature review. Tsunami simulations have been performed according to these defined 30 earthquake scenarios by tsunami numerical code NAMI DANCE (NAMIDANCE, 2011) which solves Nonlinear Shallow Water Equations (NLSWE) using leap-frog scheme. For each earthquake scenario, tsunami hydrodynamic parameters, mainly maximum water surface elevations, arrival time of first wave and maximum wave, and water level fluctuations were calculated at 1333 synthetic gauge points meticulously selected along the coasts of Marmara Sea. The overall simulation results indicate that maximum expected wave heights due to these earthquake scenarios are between 1 m and 2 m and even more than 2 m at some locations along Marmara coasts, such as Kadikoy, Halic and Silivri coasts in Istanbul and Bayramdere and Kursunlu districts along the coasts of Bursa province. The estimated maximum water levels at Bostanci, Pendik and Buyukada coasts in Istanbul, Cinarcik and Bandirma towns and at the entrance of Izmit Bay would reach up to 2 m. Tekirdag coasts and Buyuk Cekmece and Bakirkoy coasts in Istanbul and Yalova coasts would experience maximum tsunami wave amplitudes around 1.5 m. The waves reach up to 1 m at Izmit and Gemlik Bays, Erdek Peninsula and Marmara Island. The overwiew of the results reveal that higher historical tsunami wave heights observed in Marmara Sea cannot be explained by only earthquake-generated tsunamis. Therefore, there is strong agreement on considering submarine landslides as the primary tsunami hazard component in the Marmara Sea as experienced during history and expected in the future.