This study presents a real-time technology for detecting and classifying geophysical events using acoustic signals. Initially developed for tsunami monitoring, the system integrates advanced computational models and machine learning algorithms to process acoustic data and extract event characteristics, including location, magnitude, and fault dynamics [1]. The methodology facilitates real-time mapping of risk areas and event trajectories, providing timely insights critical for effective response strategies. Validation against historical events demonstrates robust performance, with global-scale analyses completed within seconds on standard multi-core machines.
An additional feature of this technology is its potential applicability to a wider range of geophysical events, such as underwater explosions and other seismic activities. Its flexibility positions it as a complementary tool for existing warning frameworks, with possible relevance for organisations like the CTBTO in future monitoring efforts.
By addressing key challenges such as false alarms and response delays, this system contributes to improving global event monitoring and enhancing disaster preparedness. It provides a valuable resource for decision-makers aiming to mitigate risks and ensure public safety.
[1] Kadri, U., Abdolali, A., and Filimonov, M.: GREAT v1.0: Global Real-time Early Assessment of Tsunamis, Geosci. Model Dev. Discuss. [preprint], https://doi.org/10.5194/gmd-2024-139, in review, 2024.
How to cite: Kadri, U., Abdolali, A., and Filimonov, M.: Real-Time Global Assessment of Tsunami Risks Using Acoustic-Gravity Waves, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6080, https://doi.org/10.5194/egusphere-egu25-6080, 2025.
Tsunami-induced currents were detected by the Taiwan High-Frequency Radar (HFR) Network on April 3, 2024, following a magnitude 7.2 earthquake near Hualien, Taiwan. The earthquake triggered tsunamis that generated strong currents, leading to collisions between drifting vessels in two harbors along Taiwan's east coast. At Hualien Harbor, tsunami waves reached a height of 1.8 m, with near-field waves rapidly propagating along Taiwan’s eastern coastline and extending to southern Okinawa, Japan.
Since 2019, 19 HFR stations have been deployed along Taiwan's coastal regions. These stations are capable of detecting surface currents in water depths of up to 100 meters, with a minimum velocity sensitivity of 0.02 m/s. Observations revealed that tsunami-induced radial currents in Yilan Bay reached speeds of up to 0.2 m/s. Using ensemble empirical mode decomposition, three oscillation modes with periods of 8.8, 14, and 30 minutes were identified in the HFR data. The first two modes are closely linked to the continental shelf, which has an average depth of approximately 100 m and a width of 5 km. The 30-min mode, however, is more pronounced within 15 km offshore in Yilan Bay, where interactions between the northeastern Taiwan countercurrent, Kuroshio, and tidal currents likely influence these oscillations.
Understanding these dynamics is critical for the development of an integrated early warning system supported by real-time HFR monitoring. This study highlights the value of such remote sensing systems in providing crucial insights into tsunami-induced hazards in coastal environments.
Figure. Surface currents response to tsunami oscillations. 100 and 500 m water depth shown in dashed and solid lines, respectively.
How to cite: Cheng, A., Lin, L.-C., Chien, H., Chang, H. M., Lai, J. W., Yu, H. Y., Cheng, H.-Y., and Flament, P.: Tsunami Responses as Observed by Taiwan High-Frequency RadarNetwork, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5329, https://doi.org/10.5194/egusphere-egu25-5329, 2025.
Ocean-bottom observations are essential for studying earthquake and tsunami processes in the ocean. Traditionally ocean-bottom pressure gauges (PGs) were used to observe tsunamis, while recent studies have revealed that they capture geophysical phenomena across a wide period range from seconds to years. Utilizing this capability, I have analysed in-situ PG data, recorded inside the earthquake source region, to reveal the physics of massive tsunami generation. In this presentation, I introduce my recent works related to the analyses of the in-situ PG data, which provides important insights into earthquake and tsunami mechanics, particularly in the Tohoku subduction zone, in northeastern Japan.
First, I highlight advances in earthquake source modelling using dynamic pressure changes recorded by the in-situ PGs (Kubota et al. 2017; 2021). PGs can detect not only tsunamis ranging periods of ~102–103 s but also dynamic pressure changes caused by seismic waves, covering periods of 100~102 s (e.g. Filloux 1982). Applying solid-fluid coupled wave theory (e.g. Saito 2019), I developed a technique to simulate the dynamic pressure fluctuations and successfully modelled broadband in-situ PG data including both long-period tsunamis and short-period seismic components. This method integrates the spatial reliability and robustness of tsunami data with the temporal resolution of seismic data.
Next, I present a case study of the 2011 Tohoku earthquake using the in-situ PGs to explore the mechanics of its large near-trench slip (> 50 m) resulting in a devastating tsunami (Kubota et al. 2022). While the kinematics of this anomalous slip have been studied well, its driving force and underlying physics remain unresolved. Using the in-situ tsunami waveforms recorded by the PGs together with geodetic datasets, I reliably estimated the distributions of slip and shear stress release on the megathrust fault plane. The results showed the near-trench slip (> 50 m) occurred with minimal stress drop (< 3 MPa) at depths < 10 km, while large stress release (> 5 MPa) occurred deeper near the hypocenter ~15 km). This suggests the near-trench slip occurred without releasing the pre-accumulated shallow stress but was driven instead by strain energy releases in the deeper region under the free-surface effects near the trench. This implies that similar large shallow slips could occur in other subduction zones even without significant shallow strain energy accumulation but only with deeper stress release.
Seafloor pressure observations have significantly advanced tsunami propagation modelling and the evaluation of tsunami source kinematics. Integrating in-situ observations with solid-fluid coupled wave theory refines kinematic modelling and enhances understanding of tsunami generation mechanism and underlying physics. Unravelling the physics of massive tsunami generation is crucial for assessing a wide range of potential future tsunami sources, including megathrust events, tsunami earthquakes, and sequential earthquake rupture scenarios.
How to cite: Kubota, T.: Unravelling Physics of Massive Tsunami Generation Using In-Situ Ocean-Bottom Pressure Gauge Data, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3058, https://doi.org/10.5194/egusphere-egu25-3058, 2025.
Tracking changes in sea-surface height with ship-based GNSS can be used to detect tsunamis in the open ocean. In the North Eastern Atlantic and Mediterranean region case studies based on historical events show that regardless of the tsunami source (seismic, volcanic, landslide or a multi-combination), ships are likely to be in position to be among the first sensors reached. Similar results are found in the Pacific region and worldwide as ships have an excellent spatial and temporal coverage of the most active tsunamigenic zone. A network of ships, based on voluntary participation of cargo and tanker vessels could then contribute to tsunami warning, augmenting the existing systems, which are mostly constituted by tide gauges and DART. To further analyze the potential contribution of a ship-based GNSS tsunami detection network, we have implemented an automatic process that is launched for each ongoing potential tsunami event. It generates a map with a tsunami amplitude model, based on the inferred earthquake source source, and calculates the number of ships and the number of tide gauges and DART sites projected to be reached by the tsunami function of time. In all cases, ships make a significant contribution to the rapid and accurate characterization of a tsunami event as they provide observations from otherwise unsampled locations. This system could be particularly impactful in the Mediterranean and south-west Pacific regions, where many countries and islands have no direct instrumentation for tsunami detection, but the global nature of GNSS and ship routes make this technique a promising, low-cost approach, to augment tsunami detection everywhere.
How to cite: Foster, J. and Thomas, B.: Real-time contribution of ship-based GNSS contribution to tsunami warning, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9689, https://doi.org/10.5194/egusphere-egu25-9689, 2025.
This study focuses on the adaptation and enhancement of the Non-hydrostatic Multilayer version of the Tsunami-HySEA code to better model tsunamis triggered by granular flows. The aim is to optimize the code for operational hazard assessment, particularly for the Stromboli Island, and to meet the integration requirements for early warning systems for events occurring along the Sciara del Fuoco scarp.
Significant improvements have been achieved, including a 50% reduction in computational time compared to the previous version. A prototype simulation has been developed to model tsunami propagation in the Tyrrhenian Sea, initiated by a landslide at Sciara del Fuoco on Stromboli Island. This simulation employs a coupled approach: first, it models the initial minutes of the event near Stromboli using the Multilayer-HySEA code in UTM coordinates. Then, the resulting data are transferred as input to other Tsunami-HySEA codes (hydrostatic or non-hydrostatic) operating in lat-lon coordinates. This two-step process enables efficient and accurate modeling of tsunami wave propagation across the entire Tyrrhenian Sea.
Acknowledgments: This contribution was supported by the Center of Excellence for exascale in Solid Earth (ChEESE-2P) funded by the European High Performance Computing Joint Undertaking (JU) under grant agreement No 101093038 and by the grants PID2022-137637NB-C21 and PID2022-137637NB-C22 funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe”.
How to cite: Rodríguez Gálvez, J. F., Castro Díaz, M. J., Macías Sánchez, J., Lorito, S., Romano, F., de' Michieli Vitturi, M., Escalante Sánchez, C., Tadini, A., Brizuela, B., Gonzalez Vida, J. M., and Cerminara, M.: Benchmarking Non-Hydrostatic Tsunami-HySEA and Multilayer-HySEA Codes for Tsunami Hazard Assessment, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19927, https://doi.org/10.5194/egusphere-egu25-19927, 2025.
Probabilistic tsunami risk assessment is a puzzling issue due to the many uncertainties involved. Several approaches have been tried and a variety of risk metrics were used so far. A data-driven method is an alternative approach, which was tested successfully for the entire Mediterranean region and its main oceanographic basins (Triantafyllou et al., PAGEOPH, v. 180, 2023). We continue this effort by testing the approach to a set of discrete coastal spots that have historically been affected by past tsunamis. The impact metric of a tsunami is expressed in terms of tsunami intensity values, K, assigned on a 12-degree scale similar to macroseismic scales. In a coastal spot tsunami risk was calculated on the basis of the past impact data in terms of tsunami intensity. The probabilistic model adopts that the tsunami intensities observed along a stretch of coastline are continuous and independent random values, with activity rate, r, distributed according to an exponential law similar to Gutenberg-Richter law with slope b. The so-called Hard Bounds Model was followed to account for the uncertainty involved in tsunami intensity determination, implying that the real, unknown tsunami intensity is assumed to occur within fixed boundary limits. The coastline-characteristic tsunami risk parameters r, b, Kmax are estimated using a maximum likelihood estimation technique. The procedure allows utilization of the entire data set consisting not only from the complete (recent) part of tsunami catalogue but also from the highly incomplete and uncertain historical part of the catalogue including palaeotsunami data, if any. Aleatory and epistemic uncertainties in the occurrence model are approached using a mixing Poisson-gamma distribution based purely on empirical data as an alternative formalism to the classic Bayesian method. The method was applied to a series of test-sites including the cities of Rhodes, Heraklion, Aegion, Zakynthos in Greece; the Augusta bay (east Sicily) and the volcanic island of Stromboli in Italy, and Algiers in Algeria. Tsunami risk is assessed in terms of probabilities of exceedance and return periods of certain intensity values in specific time frames.
How to cite: Papadopoulos, G., Triantafyllou, I., and Kijko, A.: Data-driven probabilistic tsunami risk assessment from incomplete and uncertain historical impact records in Mediterranean coastal sites, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16090, https://doi.org/10.5194/egusphere-egu25-16090, 2025.
n recent years, increasing attention has been given to evaluating potential hazards in areas of interest for offshore activities, such as potentially triggered seismicity offshore and its cascading effects, including possibly triggered landslides and tsunamis. In the present work, carried out under the SPIN project ("Test delle Buone Pratiche per lo studio della potenziale interazione tra attività offshore e pericolosità naturali" – “Testing good practices for the study of the potential interaction between offshore activities and natural hazards”), funded by the Italian Ministery for the Environment and the Energetic Security, we present a methodology to model earthquake generated tsunamis and we apply it to two study areas: “Alto Adriatico”, on the Italian central-northern Adriatic coast (southern Emilia-Romagna, Marche and northern Abruzzo regions), and “Canale di Sicilia”, on the southern coast of Sicily around the Gulf of Gela.
The first step of the workflow consists in combining multichannel 2D and high-quality 3D seismic data, morpho-bathymetric data, instrumental seismicity records, and well data to characterize both shallow and deep tectonic features as well as active faults. Then, 3D geological and velocity models at crustal scale are built, in order to simulate the ground shaking with the identified faults with different methods, such as ShakeMap and 3D broadband ground motion simulations.
The identified faults are also considered as potential tsunamigenic sources. The tsunami generation is modelled as an initial condition problem, where the initial water displacement is determined by the coseismic displacement of the seafloor generated by a fault rupture. For a given event, we determine the magnitude of the generated earthquake from scaling laws, assuming the entire fault ruptures. Different possible slip distributions are considered. The propagation of the tsunami is computed under the shallow water approximation on a system of rectangular nested grids. Increased spatial resolution is used in areas of interest, such as harbours and industrial complexes.
In the “Alto Adriatico” area, we consider thrusting faults with magnitudes up to 6.5. However, due to the very shallow bathymetry, the tsunami simulations show modest maximum amplitudes and limited inundation andwatercourses seem not to be affected. Late sea level oscillations are observed in the Ancona harbour. In the “Canale di Sicilia” area, we consider faults with a normal mechanism, with magnitudes up to 6,0. Maximum amplitudes, while still modest, are more sensitive to changes in the location of the maximum slip portion of the fault.
How to cite: Angeli, C., Armigliato, A., Zanetti, M., Zaniboni, F., Carcano, S., Forzese, M., Lipparini, L., and Molinari, I.: Potential tsunami hazard in the central Adriatic and southern Sicilian coasts associated with offshore activities, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10778, https://doi.org/10.5194/egusphere-egu25-10778, 2025.
The Gallipoli Peninsula, located at the junction of the Marmara and Aegean Seas in northwestern Turkey, has long been of strategic importance both commercially and militarily. However, its geographical location also makes it particularly vulnerable to natural disasters originating from these adjacent seas. Bordering by the active North Anatolian Fault Zone (NAFZ) to the east, the Ganos Fault to the west, and faults associated with the Biga Peninsula to the south, the region is at a high risk of seismic events, including earthquakes and subsequent tsunamis.
Despite extensive research on the seismicity of the peninsula, there remains a significant gap in understanding its specific vulnerability to tsunamis. This study aims to address this deficiency by employing advanced numerical simulation methods, in particular the Cornell Multigrid Coupled Tsunami Model (COMCOT) and Tsunami-HySea algorithms. COMCOT, known for its use in modelling the 2004 Indian Ocean tsunami propagation in regions such as Aceh, Indonesia, and the 2006 South Java tsunami on Widarapayung Beach, simulates tsunami propagation and inundation over complex bathymetries with a focus on coastal impacts. On the other hand, tsunami-HySea, which has been employed for tsunami impact analysis in areas such as central Chile for cities such as Coquimbo and Valparaíso, offers a high-resolution, multi-layer approach to understanding tsunami dynamics, which is particularly useful for detailed inundation mapping.
We designed two earthquake scenarios for each of the submarine extensions of the NAFZ and the Ganos Fault, areas with a high likelihood of seismic activity. The simulations conducted with these algorithms indicate that the Gallipoli Peninsula faces significant tsunami risks, particularly along its western and eastern coastal settlements, challenging the common perfection of low risk. These findings highlight vulnerabilities in both infrastructure and superstructure, suggesting the need for an early warning system, public education on tsunami risks, and the identification of structural vulnerabilities. This research highlights the need for further preparedness and prevention measures as the peninsula’s population and development on the increases.
How to cite: Tülümen, E. B., Tarı, U., Aydın, İ. G., and Duru, M. A.: Assessing Tsunami Hazards via COMCOT and Tsunami-HySea Simulation Algorithms on the Gallipoli Peninsula, NW Türkiye, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16751, https://doi.org/10.5194/egusphere-egu25-16751, 2025.
Potential large earthquakes in the Manila Trench of the South China Sea cannot be ignored, and the tsunami caused by potential earthquakes will impact the southern coast of China. This study aims to quantify the probability of the nearshore tsunami height along the coast of the Guangdong–Hong Kong–Macao Greater Bay Area (GBA), which is one of the most densely populated and economically developed regions in China. An approach for probabilistic tsunami hazard assessment has been established, which comprehensively considered the uncertainty in earthquake epicenter, magnitude and focal depth through numerical simulations of more than one million potential earthquake scenarios. We have developed efficient alternative algorithms for the deep-sea propagation of tsunami waves and its nearshore amplification, making the analysis of massive tsunami scenarios a reality. Through the simulation and statistics of potential tsunamis, a tsunami wave height dataset with a spatial resolution of 0.1° covering the South China Sea was established, and a refined dataset of tsunami wave height distribution along the coast of GBA was provided.
How to cite: Niu, X. and Gao, X.: Nearshore tsunami height probability along the coast of the Guangdong–Hong Kong–Macao Greater Bay Area, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2981, https://doi.org/10.5194/egusphere-egu25-2981, 2025.
Landslide tsunami hazard analysis is associated with high uncertainty. In other words, predicting the temporal exceedance probability for a given tsunami height will involve a very large uncertainty. As such, a common hazard methodology does not exist for landslide tsunamis. Most approaches are based on scenario analysis, while the Probabilistic Tsunami Hazard Analysis (PTHA) methods are rarely employed. A reason for the lack of a streamlined approach is arguably the uncertainty, related to lack of past landslide tsunami data that can provide a statistical background for most areas across the world. Modelling procedures also need a higher degree of sophistication than for earthquake tsunamis, particularly for the subaerial landslide sources producing impact tsunamis.
The authors of this abstract have previously developed a Landslide PTHA (LPTHA) that was used for tsunami hazard mapping in Norway. It combines landslide rates derived from slope stability assessment, with an event tree analysis of landslide kinematic parameters. To model the LPTHA uncertainty, it was necessary to run thousands of simulations using a range of values for the landslide parameters with highest influence on generated tsunami (e.g. runout distance, impact velocity, and frontal area). To accomplish this, the models were simplified, and hence, a significant model uncertainty is propagated. A persistent shortcoming of this method was that the chosen probabilities, landslide parameters, and outputs from the hazard models were not tested towards observations of previous events. To constrain the uncertainties, we propose here a new method comparing run-up observations of past events with simulations based on sets of uncertain parameter values. In this presentation, we combine a block source model with a linear dispersive tsunami propagation model coupled with a non-linear shallow water inundation model. By simulating a few historical rockslide tsunami events with this procedure, we analyse how LPTHA event sets match past models. We finally analyse which parameter datasets that are presently considered most suitable for future LPTHA forecasts.
How to cite: Løvholt, F., Glimsdal, S., and Harbitz, C. B.: Calibrating Probabilistic Tsunami Hazard Analysis workflows for subaerial landslide sources, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12768, https://doi.org/10.5194/egusphere-egu25-12768, 2025.
On 17 December 2024, a strong, damaging earthquake (Mw7.3) occurred 30 km WNW of the capital of Vanuatu, Port Vila, on Efate Island. The epicentral depth of the event is at ~57 km depth in the central region of the Vanuatu Subduction Zone (VSZ). Within minutes following the main shock, a small tsunami was recorded on Port Vila coastal gauge (VANU), located at the end of Mélé Bay, along a wharf of the commercial port, with a maximum recorded amplitude of ~29 cm, below the threshold (30cm) for the Beach and Marine Threat warning category. However, it should be noted that it may have overtopped this value in other local locations; for example, nearby Erakor Lagoon, which often showed larger tsunami impact during past events. Later, the tsunami was recorded on the deep-ocean Bottom Pressure Recorders of the NZ DART regional network (DART NZL and NZK), and other Vanuatu gauges and in New Caledonia, including the Loyalty Islands. Apart from VANU and LENA (Lenakel, Tanna Island, ~17 cm), the records do not show maximum tsunami amplitude greater than 10 cm. There is also no evidence that the tsunami waves reached other gauge locations in the southwest Pacific region, such as Fiji, Tonga, New Zealand, and Australia. At present, we are not aware of reports of damage related to the tsunami.
In order to simulate the tsunami generation and propagation in the region to support rapid response, simple seismic rupture models were quickly designed based on available moment tensor solutions from different seismological agencies. It also used empirical laws, past events, and geologic knowledge of the region. Several models were tested, as the moment tensor solutions did not elucidate which structure within the over-arching active process of the VSZ was responsible. Interestingly, the azimuths and dip angles of the nodal planes do not fit well with the subduction interface as it is known. The simulation of the tsunami was initially performed on a limited number of nested grids to reduce the computational costs in an event response framework, focused on providing refined solutions for the Efate and Port Vila region only. Further simulations encompassed more nested grids in New Caledonia, Fiji, New Zealand, Tonga and Australia (including Norfolk Island). Comparison of the simulation results with the recorded waveforms in Vanuatu and New Caledonia show a good agreement in terms of arrival time, phase and amplitude. Modelled maximum wave amplitude maps confirmed that the tsunami did not exhibit amplitudes larger than ~30 cm and did not propagate out of the Vanuatu-New Caledonia region. More detailed simulations demonstrate that tsunami arrivals may have exceeded 40 cm locally, for example in Fatumaru Bay (northeast of Mélé Bay).
In addition to the validation of a quickly-designed rupture model for tsunami assessment within minutes of an earthquake occurrence, the strong alignment between the simulations and observations suggests that the source of this tsunami was sea-floor displacement related to oblique-normal faulting during the Mw7.3 earthquake. As of time of writing, no aftershocks have generated tsunami.
How to cite: Roger, J.: Numerical simulation of the tsunami triggered by the 17 December 2024 Port Vila, Vanuatu, Mw 7.3 earthquake, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7588, https://doi.org/10.5194/egusphere-egu25-7588, 2025.
Over a period of ~5 years from August 2018 to December 2024, the Vanuatu Subduction Zone (VSZ) has demonstrated its potential to trigger large earthquakes and threatening tsunamis, particularly in its southern section, southeast of the Loyalty Islands archipelago. Of particular interest is the region where the Loyalty Ridge collides with the Vanuatu arc, where the subduction zone’s predominant strike direction changes sharply from roughly N-S to E-W. It is within this region that three earthquakes with a moment magnitude (Mw) >= 7.5 have occurred: Mw 7.5 on 5 December 2018, Mw 7.7 on 11 February 2021 and Mw 7.7 on 19 May 2023.
The latter is of major interest for three reasons: (1) it is an outer-rise event having occurred very close to the epicentre of the 1995 Mw 7.7 earthquake, which was the largest outer-rise normal faulting event globally at that time; (2) it was followed by a large set of aftershocks including a Mw 7.1 event ~1 hour after the main shock; (3) both the main shock and the larger aftershock triggered a tsunami; and (4) diverse records of the tsunamis exists, including data from the New Zealand DART network and the recently deployed SWOT satellite.
The first tsunami had sufficiently large wave amplitude to be recorded on most gauges of the southwestern Pacific Ocean, as far as Tasmania in the southwest (~3000 km) and Fongafale to the north-east (~1900 km), although the second tsunami was barely noticeable on the deep-ocean monitoring systems (i.e., DART) and sea-level coastal gauges.
Numerical simulations of tsunami generation and propagation using COMCOT modelling code were performed with different source models to try to deduce the source characteristics, however despite the array of available finite fault sources, none were able to fit the tsunami observations fully, including deep-ocean DART locations. In addition, SWOT 2D measurements, if generally showing a good correlation with the simulation outputs, still reveal elevations quite different from the simulation, notably in terms of amplitude of the main tsunami wavefront propagating towards the northeast. Investigations of ionospheric response to the event using GNSS records highlights the existence of a secondary source associated with the main Mw 7.7 shock, which may be linked to later release of seismic energy and/or the breaking of a second rupture patch.
This presentation aims to show what is known so far, and what are the key pieces of information still missing which may help us to explain the tsunami observations induced by the 2023 earthquake.
How to cite: Roger, J., Gusman, A., Faugère, Y., Rolland, L., Hébert, H., Delouis, B., and O'Kane, A.: The 2023 South Vanuatu doublet of earthquakes and tsunamis: observations, numerical simulations and gray areas, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7571, https://doi.org/10.5194/egusphere-egu25-7571, 2025.
When looking at the history of tsunami research, considering the early efforts, two trends can be observed. Academic tsunami research was carried out in diverse disciplines with boosts after large global events, such as the 1960 Chile event that lead to the creation of warning centers in the U.S. and Japan in the Pacific, or the 2004 Indian Ocean event that had a large impact on global tsunami preparedness efforts, supported by IOC UNESCO and other global organizations. On the other hand, the engineering community in particular in the United States created building codes and formalized such hazard prevention measures.
Efforts were made to gather the scientific community and the well established series of tsunami sessions at AGU and EGU meetings is just one indication for this. The IUGG Joint Tsunami Commission formalized some of the community effort in tsunami research, and the Tsunami Society International with its International Journal Science of Tsunami Hazards has been instrumental to gather important information and progress in tsunami science.
Around 2015 it became clear that there is demand for a formal approach to an integration of scientific progress and transfer into stakeholder groups, involving social sciences, geosciences, engineering, and computational sciences. Adopting some of the probabilistic approaches from seismic hazard assessment, covering uncertainty quantification, and developing multi-scale approaches to hazard and risk analysis, communicating and applying these topics was outside of the purely scientific agenda.
The idea for a Global Tsunami Model (GTM) entity was born, borrowed from the Global Earthquake Model (GEM) foundation. Further discussions within the community at several international meetings finally led to the idea of applying for a COST Action, funded through the European Cooperation in Science and Technology (COST). The COST Action AGITHAR was then instrumental in further developing and forming a basis for an entity supporting the ideas mentioned before. Accompanied by successful European Research Council funded projects related to tsunami hazard and risk assessment a portfolio of products and services could be developed. A further one-year funding from COST for sustaining the efforts of AGITHAR, finally led to the inauguration of the Global Tsunami Model Association, a registered association under German legislation.
In this presentation we announce GTM Association and invite the global community to become part of this initiative. The presentation will give a brief overview of the history of GTM, will introduce the vision and mission of the association, as well as outline the governing structure. We present the assets as well as our ideas on a sustained business model with a variety of development paths open to the community. While much of the development took place in the European context so far due to funding opportunities, GTM is global and will extend internationally. GTM is committed to serve the scientific community, stakeholder groups as well as the general society by coordination, knowledge transfer, and scientific progress as a non-for-profit organization.
How to cite: Behrens, J., Babeyko, A., Baptista, M. A., Denamiel, C., González Vida, J. M., Hancilar, U., Jalayer, F., Lorito, S., Løvholt, F., Macias, J., Murphy, S., Özer Sözdinler, C., Ragu Ramalingam, N., Romano, F., Rudloff, A., Selva, J., Volpe, M., and Kanoglu, U.: Announcing the Global Tsunami Model Association, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5074, https://doi.org/10.5194/egusphere-egu25-5074, 2025.
The 2022 Hunga Tonga–Hunga Ha'apai (HTHH) eruption and tsunami demonstrated the need to be better able to model tsunamis generated via multiple source mechanisms and with impact at scales from local, to regional, and global. There have however been many other examples of complex geophysical events that generate tsunamis either by a multiplicity of sources or cascades of events: e.g. the 2018 Palu event, the Aysen fjord tsunamis in 2008, the Flores Island tsunami in 1992, and the 1964 Prince Willams Sound tsunami. High Performance Computing (HPC) is necessary to be able to provide the necessary temporal and spatial resolution needed for modelling the multiple physics sources and tsunami propagation and inundation for events such as HTHH. Within the ChEESE-2P project funded by EuroHPC, a workflow is presently being developed to simulate the impact of complex tsunamigenic events in both near and far fields leveraging HPC resources. A set of numerical models optimized for HPC are coupled within the workflow: SeisSol (for modelling earthquake sources, tsunamigenesis, and acoustic coupling), ExaHyPE (for modelling gravitational flows and water wave propagation), MultiLayer-HySEA (for modelling near-field tsunami generation), Meteo-HySEA (for modelling tsunami generation driven by atmospheric waves), and Tsunami-HySEA (for modelling regional and global tsunami propagation and inundation). In this presentation, we outline the workflow, including the individual application modelling components, their coupling, and the plan to couple the models together for a joint future simulation for the HTHH event using the workflow.
How to cite: Gibbons, S. J., Bader, M., Denamiel, C. L., Díaz, M. J. C., Gabriel, A.-A., González del Pino, A., Lorito, S., Macías Sánchez, J., Romano, F., Storrøsten, E. B., Ulrich, T., Wille, M., and Løvholt, F.: A workflow for Complex Multi-Source Tsunami Modelling, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11615, https://doi.org/10.5194/egusphere-egu25-11615, 2025.
TS-GAUSS (http://ts-gauss.rm.ingv.it/) is a Virtual Access service which provides a rapid method to model tsunami propagation for a set of predefined points of interest (POIs) using a dataset of pre-calculated tsunami waveforms. This web application exploits the concept of the surface Green's functions described in Molinari et al., 2016, and consists of two consequent steps: (1) simulation of the initial tsunami conditions for an arbitrary seismic source and (2) linear combination of Gaussian-shaped elementary sources uniformly distributed across the Mediterranean Sea. The service supports the most common browsers (Google Chrome, Mozilla Firefox, Safari, Microsoft Edge) and no login credentials are currently required. The graphical user interface (GUI) consists of an intuitive input form to provide the seismic parameters of an arbitrary seismic source and an interactive map of the domain. The corresponding results are shown on the map and they can be easily downloaded in different formats and contents (as standalone navigable maps, static figures and/or explicit data) depending on the needs of the users. The service can represent a useful instrument for both students and the scientific community, in particular for tsunami modellers, hazard and risk analysts and for the activities connected to tsunami early warning centres. The landing page includes documentation and links to more detailed resources and a weekly report is automatically created to track the statistics of the tool’s usage. Moreover, the source code of the core module of the tool (without the web GUI) is a package of C++ routines, currently available as a gitlab repository (Babeyko et al., 2024). TS-GAUSS, in the near future, will also become one of the services hosted by the EPOS TCS Tsunami portal (tsunamidata.org).
This work has received funding from the European Union through the Geo-INQUIRE project (GA 101058518), within the Research Infrastructures Programme of Horizon Europe.
Babeyko, A., Romano, F. and Tonini, R. (2024): Tsunami simulation Green's function toolbox TS-GAUSS. GFZ Data Services. https://doi.org/10.5880/GFZ.2.5.2024.002
Molinari, I., Tonini, R., Lorito, S., Piatanesi, A., Romano, F., Melini, D., Hoechner, A., Gonzàlez Vida, J. M., Maciás, J., Castro, M. J., and de la Asunción, M.: Fast evaluation of tsunami scenarios: uncertainty assessment for a Mediterranean Sea database, Nat. Hazards Earth Syst. Sci., 16, 2593–2602, https://doi.org/10.5194/nhess-16-2593-2016, 2016.
How to cite: Vitiello, L., Babeyko, A., Bruni, S., Vallone, R., Romano, F., Tonini, R., and Lorito, S.: TS-GAUSS, a web application for rapid estimation of tsunami impact based on pre-calculated simulations in the Mediterranean Sea. , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19547, https://doi.org/10.5194/egusphere-egu25-19547, 2025.
Non-reflecting wave propagation is important for different applications of wave theory, where it is required that waves propagate over large distances without loss of energy. It has been found, that such waves exist not only in homogeneous or quasi-homogeneous media, but also in strongly inhomogeneous ones. For one-dimensional and quasi one-dimensional planar wave propagation in the ocean the corresponding solutions were found for convex bottom profiles and for a set of U- and V-shaped narrow bays and channels (Didenkulova et al. 2009, Didenkulova and Pelinovsky, 2009, 2011). However, this type of problem may also arise in the framework of cylindrical or radially symmetric waves. In geophysical applications, this corresponds to tsunami wave propagation from the meteorite fallen into the sea, or from underwater volcanic eruptions. In this work we find strongly varied sea bottom geometries, which allow for traveling wave solutions in the framework of cylindrical wave equation. Here we find two classes of non-reflecting geometries, which correspond to a bottom profiles next to (i) a radially symmetric deep sea trench and next to (ii) a volcanic island. Wave dynamics along these bottom geometries is also discussed.
The work was carried out with support from the RSF grant no. 23-77-01074.
Didenkulova, I., Pelinovsky, E., Soomere, T. Long surface wave dynamics along a convex bottom, J. Geophysical Research – Oceans, 114, C07006 (2009).
Didenkulova, I., Pelinovsky, E. Non-dispersive traveling waves in strongly inhomogeneous water channels, Physics Letters A, 373 (42), 3883-3887 (2009).
Didenkulova, I., Pelinovsky, E. Runup of tsunami waves in U-shaped bays, Pure and Applied Geophysics, 168 (6-7), 1239-1249 (2011).
How to cite: Didenkulova, I., Didenkulova, E., and Pelinovsky, E.: Non-reflecting cylindrical wave propagation in the ocean of changing depth, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12478, https://doi.org/10.5194/egusphere-egu25-12478, 2025.
The Shallow Water Equations (SWE) is widely used to simulate the ocean waves, particularly tsunami waves, given its simplicity and robustness for wide range of wave dynamics. This model is limited to the hydrostatic pressure assumption. However, in some scenarios such as landslide tsunamis and slow earthquake-generated waves, the non-hydrostatic pressure plays a crucial role. In that case, two approaches are mainly used: Boussinesq-type equations and non-hydrostatic SWE extensions. The SWE extensions can be achieved by splitting the pressure terms into hydrostatic and non-hydrostatic pressure while deriving a depth-averaged form.
We extend the work by Jeschke et al. (2017), where the quadratic pressure relation was used instead of the linear one showing equivalence to Boussinesq-type equations. This model was improved for moving-bottom generated waves and manipulated such that it can be solved by a projection method without the simplifications in the mentioned publication [Firdaus and Behrens (2024)]. Furthermore, this method also allows us to make the non-hydrostatic correction adaptively on a particular area, where the dispersion might play a significant role. In this work, we investigate such an adaptive model in simulating moving bottom-generated waves. We compare both global and local correction simulations with measured data along with their computational time. It can be shown that we can achieve a similar accuracy with lower computational effort.
References:
- Jeschke, A., Pedersen, G. K., Vater, S., and Behrens, J. (2017) Depth-averaged non-hydrostatic extension for shallow water equations with quadratic vertical pressure profile: equivalence to Boussinesq-type equations. Int. J. Numer. Meth. Fluids, 84: 569–583. doi: 10.1002/fld.4361.
- Firdaus, K., Behrens, J. (2024) Non-Hydrostatic Model for Simulating Moving Bottom-Generated Waves: A Shallow Water Extension with Quadratic Vertical Pressure Profile. *arXiv*. https://arxiv.org/abs/2410.23707.
How to cite: Firdaus, K. and Behrens, J.: Adaptive Non-Hydrostatic Model for Moving Bottom-Generated Waves, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3189, https://doi.org/10.5194/egusphere-egu25-3189, 2025.
Volcanic islands are prone to massive landslides and flank collapses, which can trigger tsunamis with devastating consequences. A notable example is the 2018 Anak Krakatau tsunami, which resulted in the loss of over 400 lives. While some of the energy from such landslides generates far-field tsunamis that propagate over large distances, a significant portion creates trapped waves that travel around the island. These trapped waves, unaffected by geometric spreading, can reach distant coastal areas on the same island, potentially causing severe localized damage.
Although numerous numerical studies have explored landslide-generated tsunamis in the context of conical islands, analytical studies that delve into the underlying physics of the phenomenon remain limited. Recent research in fluid mechanics has yet to analytically determine the discrete frequencies of trapped and radiating waves. Accurate calculation of the discrete frequency spectrum of trapped wavefields is crucial for assessing coastal hazards. In this study, we present a comprehensive analytical solution for the radiating and trapped wavefields generated by landslide sources with varying time histories on conical island flanks.
How to cite: ozeren, M. S. and Postacioglu, N.: Triggereing of radiating landslide Tsunami modes around conical islands, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3277, https://doi.org/10.5194/egusphere-egu25-3277, 2025.
Impulse waves generated by landslide differ from earthquake tsunamis in several aspects, as their generation mechanism as well as their length scale of propagation are not the same. In particular, the wave amplitude can be significant upon generation and may subsequently induce a substantial run-up in a nearby coastal area [1]. In this context, predicting the wave behavior after impact is of crucial interest. To have a better global understanding of this phenomenon, many studies have been devoted to its modelling, with a large variety of approaches, either experimental, numerical or field data investigations (see [2] for a detailed review). Yet, [2] suggest that the physical understanding of the phenomenon remains partial, even though numerous studies have been conducted over the last two decades. It appears then essential to better understand the mechanisms involved during the generation of such a wave in order to quantify the potential hazard it may represent. In our study, the phenomenon is modelled by a 2D-experimental setup using a steady and accelerated granular flow as a forcing wave generator. The study specifically focuses on the coupling between the granular flow and the wave, which is shown to be highly complex. In particular, the granular flow impact and its dynamics underwater can influence both the wave generation and its dynamics toward a propagation phase. The study of the wave-granular coupling during the generation phase leads to an empirical fit of the wave maximum amplitude as a function of a new dimensionless number based on two different Froude numbers, characterising both the impact properties and the granular flow propagation [3]. These new results allow to propose simple models including different finite time generation processes onto linear wave propagations, which are tested and compared to the experimental results.
[1] Fritz, H. M., Mohammed, F., & Yoo, J. Lituya bay landslide impact generated mega-tsunami 50th anniversary., Tsunami Science Four Years after the 2004 Indian Ocean Tsunami: Part II: Observation and Data Analysis, 153-175 (2009).
[2] Heller, V. & Ruffini, G. A critical review about generic subaerial landslide- tsunami experiments and options for a needed step change., Earth-Science Reviews. 242, 104459 (2023).
[3] Darvenne, A., Viroulet, S. & Lacaze, L. Physical model of landslide-generated impulse waves: experimental investigation of the wave-granular flow coupling., Journal of Geophysical Researches: Ocean., Journal of Geophysical Research: Oceans, 129(9) (2024).
How to cite: Lacaze, L., Darvenne, A., and Viroulet, S.: Impulse wave generated by landslide: investigation of the wave-granular flow coupling, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18069, https://doi.org/10.5194/egusphere-egu25-18069, 2025.
After the explosive eruption of the Hunga Tonga–Hunga Ha’apai volcano in January 2022, the generation of tsunamis driven by atmospheric acoustic-gravity waves, including the Lamb waves, has been intensively studied by the geoscientific community, resulting in hundreds of published articles since the eruption. These rare events, stemming from catastrophic volcanic eruptions, have the potential to generate surges reaching 1 to 10 m along more than 7% of the world’s coastlines. Despite their global hazard potential, probabilistic models that effectively capture the uncertainty of these events remain underdeveloped. Here, we lay the foundations of a new multidisciplinary field of research dedicated to the study of these acoustic meteotsunami events. Our work includes implementing stochastic meteotsunami surge models for 7 different volcanoes and for each of the most populated and/or endangered coastal cities in the world. We derive planetary meteotsunami surge hazards using a surrogate model approach which has already proven effective in providing fast and reliable forecasts in geosciences. As volcanic eruptions occur at the geological scale, we build these models through the numerical reproduction of all potential events with thousands of high-fidelity simulations accounting for three main sources of uncertainty: amplitude, wavelength and dissipation of the Lamb waves. Following this approach, our aim is twofold: first, to advance understanding of ocean dynamics during acoustically-driven events through in-depth analyses of the numerical simulations; and second, to enhance global coastal safety by integrating the surrogate models within existing early warning systems and providing actionable surge forecasts in the aftermath of volcanic eruptions.
How to cite: Denamiel, C., Esposti Ongaro, T., and Huan, X.: Silent Threat: Predicting Acoustic Meteotsunami Global Hazards, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7820, https://doi.org/10.5194/egusphere-egu25-7820, 2025.
The 2022 eruption of the Hunga Tonga-Hunga Ha’apai volcano, located near the Tonga Islands, resulted in a massive volcanic explosion that triggered a transoceanic atmospheric tsunami. While satisfactory scientific analyses have been conducted regarding the minor tsunami generated by the initial atmospheric pressure wave, there is still insufficient scientific discussion regarding the amplification of tsunami wave amplitudes between the first pressure tsunami wave and the volcanic gravity tsunami wave.
This study focuses on the analysis and discussion of the second group of large-amplitude tsunami waves, in addition to the first pressure wave. Our findings indicate that the oceanic disturbances in this event were primarily driven by atmospheric shock waves traveling at different velocities. The most prominent atmospheric pressure fluctuation, which reached the observation stations fastest, was a Lamb wave with an amplitude of approximately 2 hPa and a wave speed of around 308 m/s. When compared to tide gauge records from Taiwan, the sea level variation caused by this pressure wave was only about 2–5 cm. However, the sea level oscillation did not decrease but instead amplified approximately five times 2–4 hours after the first pressure wave. Through a series of numerical simulations and analyses, we found that the first pressure wave was insufficient to cause the sustained amplification of the tsunami wave amplitude. Given that atmospheric pressure propagation is much faster than that of tsunami waves, Proudman resonance is not the factor responsible for the amplification of the tsunami wave amplitude.
In this study, simulations and analyses were performed for the Taiwan region using atmospheric pressure data from the Central Weather Administration (CWA) and the COMCOT tsunami model. The pressure stations in Taiwan recorded the arrival of the first pressure wave followed by secondary, tertiary, and quaternary atmospheric gravity waves with speeds of approximately 280 m/s, 250 m/s, and 220 m/s, respectively, about 4 hours after the initial wave. By using atmospheric observational data to construct a linear model for the atmospheric gravity waves, we successfully reproduced the observed phenomenon of tsunami wave amplification approximately five times. The simulation results showed a high degree of agreement with the observed amplitude and period. The tsunami propagation simulations revealed that the amplification was caused by Proudman resonance between the second, third, and fourth atmospheric gravity waves, following the Lamb wave, and oceanic gravity waves. This effect caused the slower Pekeris wave to propagate through the deep western Pacific, significantly increasing the tsunami wave amplitude.
How to cite: Wu, T.-R., Yang, P.-Y., Lin, J.-W., and Chuang, M.-H.: A comprehensive discussion on the tsunami amplification effect of the 2022 Tonga volcanic tsunamis, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7816, https://doi.org/10.5194/egusphere-egu25-7816, 2025.
Seiches are resonant oscillations that occur when gravity waves in a basin are excited at a period coinciding with one of the periods of free oscillation of the basin. When triggered by earthquakes, seiches may influence the amplitude of tsunamis. They may also play an important role in shaping sedimentary deposits occurring during these events. Our study combines in situ monitoring (performed in the framework of EMSO-France and of Maregami Türkiye-France bilateral project) and numerical modeling to characterize seiches in the Sea of Marmara, where the North Anatolian Fault system causes large earthquakes associated with turbidite-homogenite deposits and tsunamis. Pressure sensors deployed at five different locations at the seafloor in the Sea of Marama basins recorded bursts of small amplitude oscillations (< 1 hPa) with periods ranging from 5 to 200 minutes, apparently triggered by storms. Resonance spectra were extracted by cepstrum analysis, a method commonly used in speech recognition. Observed resonance modes were characterized by their period at peak amplitude and by their log amplitude at each deployment location. Theoretical free oscillation modes were calculated as eigenvalues and eigenvectors of the shallow water equation with the best available bathymetry (99 modes were calculated with periods ranging 17 to 183 minutes). These provide a better match of observed resonance frequencies than the shortcut calculation of Yalciner and Pelinovki (2002), especially at long periods (> 80 minutes). However, all calculated modes involve resonances in the shallow parts of the Sea of Marmara (shelves and bays) and most have low amplitudes in the deep basins, which may hinder their detection. Thus, it has not been possible to match observed and calculated modes one by one, but some observed-calculated pairs have fitting periods and fitting spatial variations in amplitude. Of specific interest, matching modes detected at periods of about 25 minutes have large theoretical amplitudes at the Istanbul coast, which may help explain historical reports and sedimentological evidence of tsunamis.
How to cite: Henry, P., Sinan, Ö., Postacioğlu, N., Chevalier, C., Papoutsellis, C., Paté, A., Çağatay, N., Yakupoğlu, N., and Cakir, Z.: Detection of meteorologically triggered seiche oscillations, eigenanalysis, and implications for tsunami hazards in the Sea of Marmara, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2896, https://doi.org/10.5194/egusphere-egu25-2896, 2025.
Atmospherically-driven tsunamis or meteotsunamis are generated by atmospheric disturbances with steep gradients of pressure and/or wind. In recent years, meteotsunamis have received more attention from the tsunami modelling community. Although their destructive potential might be less severe than for earthquake or landslide generated tsunamis, their frequency is much higher. The two main processes driving the most extreme meteotsunami events are the offshore amplification of the ocean long-waves due to Proudman or Greenspan resonances (i.e., when the atmospheric disturbance travels at the same speed than the long-waves) and, nearshore, the amplification factor of the shelfs, bays or inlets (i.e., resonance frequency associated to the nearshore geometry). As meteotsunamis have a high dependence on the nearshore geometric characteristics, they often occur at known hotspot locations such as along the coastlines of Croatia, the Balearic Islands, Sicily, Malta, the Nagasaki Bay or the Baltic Sea. One of the highest meteotsunami waves ever witnessed (with conservative estimate of up to 6 m in height) took place in Vela Luka (Adriatic Sea, Croatia) on the 21st of June 1978.
Meteo-HySEA is a GPU accelerated code developed by EDANYA group from the University of Málaga which incorporates the atmospheric forcing together with additional terms such as the Coriolis force and the wind drag to simulate meteotsunami events. After successfully benchmarking the code to replicate laboratory experiments on Proudman resonance and a real-world test case in the Gulf of Mexico using actual topobathymetric data and synthetic pressure data, this updated version of the code introduces the capability to use multiple grids with varying resolutions in a single simulation. This enhancement provides more accurate modelling of Greenspan resonance effects and enables the computation of high-resolution meteotsunami inundation.
The Adriatic Sea was selected as an ideal starting point to showcase the reliability of Meteo-HySEA, given its extensive historical record of meteotsunami events and readily available meteotsunami data. Future efforts will focus on comparing the performance of this code with other existing tools designed for meteotsunami simulations.
Acknowledgments: This contribution was supported by the EU project “A Digital Twin for Geophysical Extremes” (DT-GEO) (No: 101058129) and by the Center of Excellence for exascale in Solid Earth (ChEESE-2P) funded by the European High Performance Computing Joint Undertaking (JU) under grant agreement No 101093038.
How to cite: González del Pino, A., Macías Sánchez, J., Castro Díaz, M., and Lumina Denamiel, C.: Meteo-HySEA: A GPU accelerated code for simulating atmospherically-driven tsunamis on real bathymetries. Evaluating the performance of the newly implemented nested grids system., EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9150, https://doi.org/10.5194/egusphere-egu25-9150, 2025.
High-frequency sea level oscillations are gaining prominence in sea level research, as advancements in technology and data collection allowed high-resolution records. Their extreme manifestations, often amplified by interactions with other strong oscillations, can trigger destructive flooding events worldwide, emphasizing the need for in-depth studies of such phenomena and the development of reliable predictive tools. To tackle this, the synoptic index-based model has been designed to reconstruct and predict extreme non-seismic sea level oscillations at tsunami timescales (NSLOTTs). Initially developed for the meteotsunami hotspot Ciutadella, the model was later extended globally, with the strongest synoptic index-NSLOTT correlations observed in the Mediterranean Sea, where NSLOTTs contribute up to 50% of the total sea-level range.
The baseline model, built using ERA5 reanalysis with synoptic variables previously identified as relevant for known NSLOTT hotspots, was subjected to modifications in its configuration in order to evaluate adaptability and robustness in forecasting and detecting extreme NSLOTT events. These modifications included testing alternative reanalysis products, different synoptic variables, and training/testing datasets. Additionally, the impact of changes in NSLOTT series—such as altered temporal resolution, amount of data gaps, and series length—was assessed. Results reveal that stations with higher baseline performance consistently maintain their skill across different model configurations, though their performance variability is greater compared to stations with lower baseline performance. Stations along the eastern Adriatic Sea exhibited the highest performance, highlighting the suitability of the model for this region of Mediterranean. Overall, the model demonstrates higher success in forecasting extreme events than in their detection. These findings offer valuable insights for optimizing model configurations and enhancing predictive capabilities, with the ultimate goal of developing reliable tools for forecasting extreme events, and consequently contributing to coastal hazard and flooding mitigation.
How to cite: Zemunik Selak, P., Vilibić, I., Denamiel, C., and Pranić, P.: Synoptic index-based model for reconstructing high-frequency sea level oscillations in the Mediterranean, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-537, https://doi.org/10.5194/egusphere-egu25-537, 2025.
Meteotsunamis, or atmospherically generated tsunamis, can generate hazardous high frequency sea level oscillations in coastal regions. The inlet of Ciutadella, located on the western coast of Menorca (Balearic Islands), is a well-documented hotspot for meteotsunamis. In late spring and summer, Ciutadella frequently experiences sea level oscillations exceeding 1 meter, and occasional events of several meters have caused significant damage to boats and harbor infrastructures. Although Ciutadella has concentrated most of the attention, other locations across the Balearic Islands and the northeastern coast of the Iberian Peninsula also experience notable meteotsunamis.
This study examines the occurrence of meteotsunamis from a regional perspective, using all the available high-resolution tide gauge data with a 1-minute sampling rate collected over the past decades. The dataset includes contributions from operational tide gauge networks managed by Puertos del Estado, SOCIB, and PortsIB, the VENOM ultra-dense research network (operated by UIB, IEO-CSIC and UPC) and an individual tide gauge maintained by Josep Pascual at l’Estartit. In total, the analysis encompasses data from 27 instruments spanning the Balearic Islands and the northeastern Iberian Peninsula, with some time series exceeding 17 years and more than 10 series exceeding a decade.
Our regional analysis focuses on four key aspects: i) to characterize meteotsunami statistics across the study area including many locations that were not analysed before; ii) the contribution of the meteotsunami frequency band (1 min-2h) to sea level extremes; iii) a comparative analysis of meteotsunami events observed at different locations; and iv) the relationship between synoptic atmospheric patterns and meteotsunami occurrences. The findings reveal that high-frequency sea level oscillations are amplified in locations where topographic features favor the resonance of incoming meteostunami waves. While Ciutadella remains the primary hotspot, other locations such as Vilanova, Portocolom, and Port de Sóller also frequently experience significant meteotsunamis, which was not reported before. Moreover, we have found that sea level oscillations often occur simultaneously in several locations; the reason is that meteotsunamis are triggered by atmospheric disturbances associated with synoptic-scale meteorological patterns that cover a large part or the region, affecting several locations at the same time. Finally, the analysis highlights the challenges in predicting meteotsunami amplitudes. Their intensity is influenced not only by synoptic-scale atmospheric features but also by small-scale processes in the ocean and the atmosphere that are difficult to observe and predict. This complexity makes it challenging to establish robust amplitude relationships across locations or to issue accurate forecasts for the amplitude of meteotsunami events.
How to cite: Villalonga, J., Pascual, J., Puigdefàbregas, J., Gomis, D., and Jordà, G.: Meteotsunamis in the Western mediterranean: regional analysis from high frequency sea level observations, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8648, https://doi.org/10.5194/egusphere-egu25-8648, 2025.
The Adriatic Sea is prone to meteotsunamis, with an exceptionally strong event (wave height > 2 m) observed 1-2 times per decade, and moderate events (wave height > 1 m) once every 1-2 years. Adriatic Sea meteotsunamis occur at many locations along the mainland and, more often, islands. The goal of this research is to determine potential meteotsunami risk along the Adriatic Sea coast. The risk estimate is based on numerical modeling of maximum wave heights in dependance on speed and direction of air pressure disturbances. The modeling results are then combined with the ERA5 reanalysis over the past 30 years to determine how often suitable, previously determined, synoptic conditions for meteotsunamis present over the area. Based on both the sea modeling and atmospheric reanalysis, a meteotsunami hazard level is associated with each point of the Adriatic Sea coast, and the results are shown on a detailed map.
How to cite: Bubalo, M. and Šepić, J.: Categorization of the Adriatic Sea coast based on meteotsunami hazard level, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1782, https://doi.org/10.5194/egusphere-egu25-1782, 2025.
Tsunami Early Warning Systems (TEWS) require rapid data collection, transmission and interrogation to ensure accurate and effective warnings are distributed to the public and critical infrastructure through tsunami warning centres.
MSM Ocean and Sonardyne formed a partnership to produce commercial-off-the-shelf (COTS) systems using standardised equipment. Modular, dual-redundant and field-proven systems provide operators with cost-effective, reliable and flexible deployment options with familiar existing support structures.
Each individual systems consists of a Sonardyne Bottom Pressure Recorder (BPR) and acoustic communication link to an MSM surface buoy with satellite communications to an onshore data centre. Onboard data processing reduces communication latency and therefore increases warning times. Predictable and infrequent maintenance schedules ensure these systems have high MTBF and low downtime, presenting less risk to the public.
We present case studies of current and planned TEWS in the Pacific and Mediterranean with associated tsunami events for context.
The Oceanographic Institute of the Navy (INOCAR, Ecuador) operates two arrays of Sonardyne-MSM TEWS systems located ~100km off the mainland and Galapagos Islands respectively. These arrays routinely detect sea surface height disturbances caused by events throughout the Pacific including earthquakes and volcanic eruptions.
Alerts were issued less than 60 seconds from initial seafloor BPR detection following the 2022 Hunga Tonga volcanic eruption and 2021 Mw 8.1 Kermadec Islands earthquake. At typical offshore tsunami velocities, extensive warnings and (crucially) responses to those warnings are possible with the geographic distribution of the TEWS array.
The National Institute for Geophysics and Volcanology (INGV, Italy) will install a TEWS array in the Ionian Sea in 2025 with spare units on land to achieve minimum downtime during planned maintenance in collaboration with MSM and Sonardyne. This is a key benefit of a cost-efficient and uncomplicated COTS solution.
Integration of buoys into a pre-existing network requires location optimisation to achieve maximum warning times. In this case, INGV has calculated a pair of buoy locations by minimising the cost function (maximising warning time) of several parameters including known tsunamigenic sources, associated tsunami spatial and temporal evolution, the severity and probability of such events and the existing contributions from coastal tide gauges to any alerts. The addition of offshore Sonardyne BPRs, with an acoustic link to MSM surface buoys is far more cost efficient than proposed cabled solutions.
Combining pre-existing and reliable infrastructure with additional new offshore equipment provides both the Pacific and Mediterranean coastlines with a significant increase in warning times and data availability.
How to cite: Reis, W., Zanette, C., and Rodriguez, P.: Field-proven experience of Tsunami Early Warning Systems (TEWS) in the Pacific and Future Arrays in the Mediterranean: Increased Warning Times and Data Availability, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18987, https://doi.org/10.5194/egusphere-egu25-18987, 2025.
Tsunamis are one of the most devastating natural hazards, with the potential to cause extensive loss of life, property damage and socioeconomic disruptions. Developing robust and accurate early warning systems is critical to mitigating these impacts. In this study, a neural network-based early warning system is proposed to predict tsunami wave heights nearshore, focusing on the Vancouver Island area on the western coast of Canada.
The Vancouver Island region, which is extremely susceptible to tsunami hazards because of its closeness to the Cascadia Subduction Zone, is the area used to generate the synthetic data. In tsunami research, synthetic data are essential because they enable the investigation of a variety of possible earthquake and tsunami scenarios, including uncommon but highly consequential occurrences. The dataset, which contains 5000 simulation scenarios, used includes parameters such as fault slip parameters, bathymetry, hypocenter position, and earthquake magnitude, as well as the related tsunami wave heights at particular nearshore locations. The parameters used to train the model are the maximum wave heights off shore at different stations and the parameter that the model is trained to predict is the maximum wave height near shore in different depth zones (0 m, 5 m, 10 m, and 100 m).
The neural network architecture was designed to model the nonlinear relationships between input parameters (maximum wave heights off shore at different stations) and resulting tsunami wave heights (near shore at different depths). By training, validating, and testing the neural network, the model demonstrated a high level of accuracy in predicting wave heights nearshore. The performance metrics, including mean absolute error and correlation coefficients, indicate that the neural network effectively captures the complexities of tsunami wave dynamics, making it suitable for early warning applications. According to the results, the neural network can accurately forecast tsunami heights close to shore, facilitating prompt evacuation preparation and disaster relief. This method is a major improvement over conventional physics-based models, which frequently demand a large amount of time and resources, by providing a computationally effective and scalable solution. Overall, this study demonstrates how machine learning, and in particular neural networks, might improve early warning systems for tsunamis.
How to cite: Chamatidis, I., Istrati, D., Goda, K., and Lagaros, N. D.: Early Warning Tsunami Prediction Using Neural Networks: A Case Study in Vancouver Island, Canada, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20668, https://doi.org/10.5194/egusphere-egu25-20668, 2025.
Tsunamis are among the most devastating and infrequent natural phenomena, capable of causing immense loss of life and property in coastal regions. While predicting the occurrence of tsunamis remains challenging, communities can take proactive steps to mitigate their impact. Local, national, and intergovernmental initiatives aim to provide a legal framework for strengthening community preparedness through a comprehensive approach that includes measures ranging from tsunami hazard and exposure assessments, generating evacuation maps, installing corresponding signage, and promoting education and capacity building of local stakeholders and population. It also involves the establishment of Standard Operating Procedures (SOPs) to ensure a timely and effective end-to-end tsunami warning communication chain. This study presents an overview of the recent significant progress in tsunami preparedness across countries bordering the Mediterranean and North East Atlantic coasts.
How to cite: Daskalaki, E., Aguirre Ayerbe, I., Baptista, M. A., Amato, A., Cambaz, M. D., Charalampakis, M., Cugliari, L., El-Gharabawy, S. M., Hamouda, A., Hebert, H., Kalligeris, N., Cantavella Nadal, J. V., Meral Özel, N., Péroche, M., and Yalciner, A. C.: Recent Developments in Tsunami Preparedness in the Northeast Atlantic and Mediterranean Region: Challenges, Strengths, and Weaknesses , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16032, https://doi.org/10.5194/egusphere-egu25-16032, 2025.
The villages of Palabuhan Ratu and Jayanti are identified as areas with a high tsunami hazard level due to their proximity to the South Java megathrust zone, which is estimated to have the potential to generate an earthquake of up to M9.1. Furthermore, their location within Palabuhan Ratu Bay could amplify tsunami wave heights, exacerbating the potential impact. Palabuhan Ratu is a densely populated area renowned for its beaches. Jayanti hosts critical infrastructure, including the Palabuhan Ratu Steam Power Plant, which significantly contributes to the electricity supply for Java and Bali. These factors collectively increase the vulnerability of these villages to tsunami-related risks. In this study, we present an evacuation model that combines simulation with tsunami scenarios, as well as causality and evacuation estimations. We used COMCOT to model tsunami propagation, run-up, and inundation, and TUNAMI-EVAC1 for agent based modeling to simulate community behavior during tsunami evacuations. The tsunami and agent based modeling results indicate a flow depth of up to 31 meters with an arrival time of 21 minutes, and a fatality impact of 39% of the total population of both villages if the community understands the location of evacuation sites, rising significantly to 57% if they do not know
How to cite: Weniza, W., Pandadaran, S. H., Anggraini, S., Hidayanti, H., Haryanto, F. T., Maimuna, A. K., Aristy, S. A., Rudianto, R., Yatimantoro, T., Apriani, M., Kriswinarso, T., Aditya, G. P., Setiawati, E. E., Panjaitan, O. D., Riama, N. F., and Daryono, D.: Agent-Based Modeling for Palabuhan Ratu and Jayanti Villages in Response to a South Java Megathrust Earthquake-Tsunami (M9.1): An Integrated Model of Tsunami Hazard and Human Response, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14661, https://doi.org/10.5194/egusphere-egu25-14661, 2025.
NEAM-COMMITMENT is a two-year project funded by the European Commission’s DG-ECHO, starting in 2025. We will present an overview of the project, its expected outcomes, and its synergies with previous/ongoing projects and initiatives.
The project aims to support improved tsunami risk management and planning in the North-Eastern Atlantic, Mediterranean and connected seas (NEAM) region. The project endeavors to primarily contribute to two key components of tsunami risk governance: (1) capacity building through tsunami hazard assessment and mapping at the national scale, and (2) improved tsunami evacuation planning at the local level through a novel multi-hazard approach. The project capitalizes on past and ongoing projects and initiatives, e.g., TSUMAPS-NEAM, CoastWAVE, EPOS TCS Tsunami and Global Tsunami Model, among others, while investing in cross-border knowledge-sharing through an extensive scientific and emergency management partnership, with 13 partner institutions from four NEAM countries. This will strengthen the cooperation among NEAM Member States and the Union Civil Protection Mechanism (UCPM), to ultimately enhance tsunami preparedness for effective response within the framework of the NEAM Tsunami Warning System coordinated by UNESCO-IOC.
The project’s first objective is to develop national tsunami inundation maps in Cyprus, Greece and Spain through a methodology previously used to produce tsunami inundation maps for evacuation planning in Italy. The tsunami inundation mapping methodology will utilize the NEAM probabilistic tsunami hazard model offshore inputs (NEAMTHM18; Basili et al., 2021, Front. Earth Sci.) to infer the national-scale inundation zones across large stretches of coastline in Cyprus, Spain, and Greece, using a GIS-based approach (Tonini et al., 2021, Front. Earth Sci.). The second objective addresses the need for a multi-hazard approach for effective tsunami evacuation management at the local level to complement existing tsunami evacuation management guidelines (e.g., UNESCO-IOC, Manuals and Guides 82). The proposed new approach focuses on multi-hazard cascading effects concerning tsunami evacuation management and will be tested in local pilot sites in Greece and Italy, considering the hazards of earthquake+tsunami and volcanic activity+tsunami in each pilot site, respectively.
The project objectives will be achieved through science-informed, participatory decision-making, enabling decision-makers to take ownership of the products and maximize implementation effectiveness. The methodological approach draws valuable experience from a recent cross-border collaboration on tsunami hazard and evacuation mapping for the city of Larnaca, Cyprus (Aguirre Ayerbe et al., 2025, EGU Abstract), implemented within the framework of the UNESCO-IOC CoastWAVE project, also funded by DG-ECHO. In addition to the products that will be developed for the four countries, open-access guidelines and tools will be developed to document the methodologies implemented for creating tsunami inundation maps at the national level and local tsunami evacuation maps considering multi-hazard cascading effects, to contribute to improved tsunami risk management and planning in the NEAM region and beyond. Finally, the release of Open Geospatial Consortium (OGC) Web Services will enhance compliance with FAIR (Findable, Accessible, Interoperable, and Reusable) principles for mapping products and allow support for implementing the EPOS TCS Tsunami.
How to cite: Charalampakis, M., Kalligeris, N., Graziani, L., Aguirre Ayerbe, I., Di Manna, P., Silva, V., Macias, J., Russo, D., Synolakis, C. E., Antonakos, A., Pilidou, S., D'Angelo, L., and González González, C. and the NEAM-COMMITMENT project team: The NEAM-COMMITMENT EU project aiming to support improved tsunami risk management and planning in the NEAM region, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15831, https://doi.org/10.5194/egusphere-egu25-15831, 2025.
Tsunami preparedness strategies are essential in tsunami risk governance and management due to three main factors that characterize these phenomena: they may have great devastating potential, they are unpredictable until an earthquake occurs, and they move extremely fast. Preparedness strategies are prospective measures that should be planned based on risk understanding in a pre-event phase, to best identify the emerging needs. Among them, tsunami evacuation planning is one of the most relevant strategies, especially in terms of protecting lives.
Tsunami evacuation mapping constitutes the basis of an evacuation planning process. Maps must be useful for both emergency managers and population. For that reason, they must be scientifically robust, detailed, and comprehensive, but also easy to understand and attractive. Beyond their main purpose, they are also a powerful tool for public awareness and communication activities.
The methodology developed in this study to elaborate tsunami evacuation mapping is based on the proposed concept of “science-informed participatory decision-making” that has been applied in Chipiona (Spain), and Larnaca (Cyprus). “Science-informed” means that the scientific community provides the hazard and evacuation approaches, models, and their implementation within the study area. “Participatory decision-making" refers to the active involvement of all relevant actors in the discussions, analysis, problem-solving, and decision-making to collaboratively develop the final tsunami evacuation maps. Involved actors include all public, private, academic, and civil association personnel that may be related to tsunami risk management and planning. Decision-makers and stakeholders were extensively involved in the process of understanding and translating the information from science to risk management.
Tsunami hazard assessment and the analysis of the elements of the evacuation strategy constitute two main steps for the elaboration of tsunami evacuation maps. In the case of Larnaca, a Seismic Probabilistic Tsunami Hazard Assessment (S-PTHA) approach was applied. As a result, a series of hazard zones were obtained, corresponding to different return periods and percentiles of uncertainty. The first decision asked to be made by the competent decision-makers was to select the tsunami hazard zone to be used for evacuation planning and mapping. This process was facilitated through a dedicated participatory workshop in which the concepts and methods applied (such as PTHA, average return period, and uncertainty) were explained. Additionally, the implications of selecting different tsunami hazard zones for emergency and evacuation planning (including exposed areas, population, and buildings critical for evacuation) were discussed, and stakeholder's perception and concerns were analysed and addressed.
Then, preliminary tsunami evacuation maps were developed based on a least-cost distance model to determine the optimal evacuation routes leading from any point inside the designated tsunami hazard zone to a series of assembly areas, previously identified through a multi-criteria approach. Subsequently, an additional participatory workshop and field visits were carried out with key stakeholders to identify well-known places and landmarks and potential evacuation barriers to finally validate the preliminary evacuation routes and assembly areas. Local knowledge provided by stakeholders effectively contributed to ensure the understanding and usefulness of the mapping end products for local emergency/risk managers and the community.
How to cite: Aguirre-Ayerbe, I., Kalligeris, N., Lorito, S., Merino González-Pardo, M., Charalampakis, M., Romano, F., Pilidou, S., Volpe, M., Brizuela, B., Di Manna, P., Papadimitriou, N., Tonini, R., Dimitriadis, I., and Melis, N.: Tsunami evacuation mapping co-design through science-informed and participatory decision making, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15748, https://doi.org/10.5194/egusphere-egu25-15748, 2025.
The eastern coast of Sicily, including Catania’s harbor and the tourist beach, is highly vulnerable to tsunami hazards, with a history of major events such as the January 11th, 1693 earthquake. Due to its geographic location and the region’s seismic activity, Catania remains at significant risk of similar catastrophic events. Evacuation is widely recognized as the most effective means of saving lives in an imminent tsunami event.
The Catania coastal area is densely populated and there is a significant proportion of elderly people among the residents who may face greater difficulties during evacuation. Moreover, there is a significant seasonal variation in the population since small coastal towns host many tourists during spring and summer.
In this study, we propose a probabilistic simulation-based framework for evacuation modelling. In the framework, we use Agent-Based Modeling (ABM) to develop a high-resolution digital model of the evacuation environment, including the location of people, residences, roads, and the shelters that are defined in the advisory/watch tsunami evacuation maps designed for Italian coasts (Tonini et al. 2021). We have modeled human behaviors using data collected from questionnaires and other open-source statistical databases. The ABM model simulates human behavior in response to 92 detailed tsunami inundation scenarios derived from Probabilistic Tsunami Hazard Analysis (PTHA) results (Gibbons et al., 2020).
The probability of safe evacuation is assessed for various scenarios such as daytime or nighttime exposure and the presence or absence of a tsunami early warning. This assessment is evaluated using a Monte Carlo simulation workflow, incorporating all uncertain modeling parameters. These parameters range from tsunami source characteristics (e.g., magnitude and slip) to various human response factors influenced by different behavioral patterns, such as immediate escape, freezing, or seeking information as well as choices like deciding between driving a car or walking. The model incorporates different types of agents to capture the complexity of human behavior. These agents include residents, both individuals and families across various age groups, and tourists, each characterized by distinct response patterns and decision-making processes. The probabilistic evacuation modeling results are derived by sampling the agents’ response parameters, such as individual velocity and response delays, to account for variability while maintaining computational feasibility. Preliminary results from selected scenarios with simple human behavior show that tsunami scenario parameters such as magnitude and tsunami impact (e.g., flow depth), can significantly influence the probability of safe evacuation.
How to cite: Soltani, S., Jalayer, F., Dugdale, J., Volpe, M., Lorito, S., and Ebrahimian, H.: Probabilistic Coastal Tsunami Evacuation Modelling Using Agent-based Modelling in Catania, Italy, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17105, https://doi.org/10.5194/egusphere-egu25-17105, 2025.
The 30 October 2020 Aegean Sea tsunami was triggered by an Mw 7.0 earthquake at a depth of ~15 km, which occurred in Kusadasi Bay, north of Samos Island, Greece. The moderate tsunami primarily impacted the central Aegean coast of Turkiye and the northern coast of Samos Island, Greece, with a maximum runup of ~3.8 m observed in Akarca, Izmir, Turkiye. The tsunami resulted in one fatality and several injuries in Turkiye as well as destructive effects on marine vessels, particularly in two locations, Sigacik and Akarca in Izmir Province. In Sigacik Teos Marina, more than 300 vessels experienced varying levels of damage, whereas in Akarca Fishing Shelter, all floating piers were destroyed, and more than 20 vessels were highly damaged. Despite its adverse effects, the 30 October 2020 event provided significant data on damaged marine vessels serving as a key resource for developing tsunami fragility functions in the Aegean Sea.
In this study, we aim to evaluate the potential impacts and damages induced by tsunamis on marine vessels in ports, marinas, and fishing shelters by establishing correlations between tsunami parameters and their effects through the development of fragility curves and loss functions. We focus on marine vessel damage resulting from strong currents and water level fluctuations caused by the 30 October 2020 tsunami. Pre- and post-tsunami satellite imagery of Sigacik Teos Marina and Akarca Fishing Shelter was used to document vessel characteristics and evaluate the extent of damage. High-resolution numerical modeling was employed to compute tsunami hydrodynamic parameters and correlate them with observed vessel damages via regression analysis. Model validation is conducted using simulation results obtained from three distinct seismic sources available in the literature and by comparing the model results against field observations. We present the tsunami parameters in the affected ports and the resulting fragility curves for marine vessels, which reveal the relationship between vessel characteristics and the forces exerted during the tsunami. The findings provide insights into the key factors contributing tsunami-induced vessel damage, supporting efforts to enhance the resilience of coastal infrastructure and marine operations against future tsunami events in the Aegean Sea.
How to cite: Dogan, G. G., Ozturk, E. A., Yalciner, A. C., Ozbahceci, B., and Suppasri, A.: Assessment of Tsunami Damage on Marine Vessels in the Aegean Sea Ports Following the 30 October 2020 Tsunami, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19770, https://doi.org/10.5194/egusphere-egu25-19770, 2025.
Climate and non-climate events have placed unprecedented pressure on the built environment and human communities, resulting in significant damage and fatalities in recent years. Notable examples include the 2023 Storm Ciarán in the UK (Heidarzadeh et al., 2025a) and the 2024 Noto tsunami in Japan (Heidarzadeh et al., 2024). As an island nation exposed to numerous storms each year, the UK faces significant impacts from climate change compared to many other countries. Coastal defense plays a central role in the nation’s efforts to address these challenges, with approximately 18% of its coastlines currently protected by defense structures. This is part of a broader global trend, as many countries with vulnerable coastlines are prioritizing similar measures to safeguard their populations and infrastructure.
Among various coastal defense methods, revetments are the most widely used, constructed from materials such as rock, concrete, gabions, and wood. While revetments are cost-effective and utilize simple technologies, they come with drawbacks, including high maintenance costs, environmental risks, and limited beach accessibility. To address these issues, it is essential to explore innovative approaches to revetment construction. A promising alternative is high-strength steel mesh mattresses, known as TECCO CELL revetment, which showed superior performance compared to rock armour in a recent study conducted by Heidarzadeh et al. (2025b). A TECCO CELL revetment involves enclosing small rocks in steel mesh mattresses, eliminating the need to transport large rocks, as required for traditional rock armor revetments. These steel meshes are highly resistant and durable, thus reducing maintenance costs. This report presents the results of the second phase of our laboratory modeling using pneumatic piston-made solitary waves. We compare the hydraulic performance of TECCO CELL revetment with that of traditional rock armor revetments. Our results indicate that the TECCO CELL system outperforms traditional rock armor in reducing wave run-up. This research is ongoing, and additional tests are planned. Results of the first phase are published in the study by Heidarzadeh et al. (2025b).
References:
Heidarzadeh, M., Šepić, J., Iwamoto, T. (2025a). Long-duration storm surges due to 2023 successive UK Storms Ciarán and Domingos: generation, field surveys, and numerical modelling. Ocean Modelling, https://doi.org/10.1016/j.ocemod.2024.102487.
Heidarzadeh, M., Sheibani, M. & Luis-Fonseca, R.J. (2025b). Coastal Storm Risk Reduction Using Steel Mesh Revetments: Field Application and Preliminary Physical Experiments. Pure Appl. Geophys. https://doi.org/10.1007/s00024-024-03621-x.
Heidarzadeh, M., Ishibe, T., Gusman, A.R., Miyazaki, H. (2024). Field surveys of tsunami runup and damage following the January 2024 Mw 7.5 Noto (Japan Sea) tsunamigenic earthquake. Ocean Engineering, 307, 118140. https://doi.org/10.1016/j.oceaneng.2024.118140.
How to cite: Heidarzadeh, M., Sheibani, M., and Luis-Fonseca, R. J.: Innovative tsunami and storm defense using high-tensile steel mesh revetments, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13166, https://doi.org/10.5194/egusphere-egu25-13166, 2025.
Tsunami fragility curves for building classes are essential tools for portfolio risk assessment of tsunami-prone regions. However, existing fragility data comprises mainly empirical fragility curves, which reflect the vulnerability of local building classes based on observed damage data. While useful, these empirical curves have limited applicability in regions with no recorded tsunami events or insufficient damage data. This highlights the need to expand the database with analytical fragility curves, particularly for areas lacking empirical damage records. To fill this gap, this study proposes a comprehensive framework for developing analytical tsunami fragility curves for building classes in tsunami-prone regions. The framework integrates simulation of tsunami time-history scenarios with random selection of case study buildings from identified tsunami hotspots. Fragility curves are then derived using Modified Cloud Analysis (MCA), which employs logarithmic regression of structural response estimated from high-fidelity finite-element modelling of structural response (e.g., demand to capacity ratios for different damage levels) versus tsunami intensity (e.g., flow depth, momentum flux) for a set of tsunami time histories. To illustrate the framework, a case study is presented focusing on the low-rise residential reinforced concrete (RC) buildings along the east coast of Sicily, Italy, within the Plain of Catania. An extensive set of tsunami inundation scenarios was simulated for the Catania Plain, which includes tsunamis generated by earthquakes in the Mediterranean Sea with following features: i) near- and far-field earthquakes; ii) crustal and subduction earthquakes; and iii) earthquakes with moment magnitudes from 6.0 to 9.0. For each scenario, locations with the most significant flow depths, i.e. tsunami hotspots, were identified and one building was then selected from each tsunami hotspot for structural simulation under tsunami loading. Details of the selected building structures were generated via the simulated design tools provided by EUCENTRE, which does automatic identification of possible structural design, considering both the variations of structural configuration and material properties. Finally, the resulting fragility curves for RC buildings were derived using the MCA approach and hierarchical fragility modelling for a 5-tier damage scale based on EMS 98 definition and with relative confidence intervals, providing valuable information of building vulnerability in that region.
How to cite: Wu, J., Jalayer, F., Ebrahimian, H., Volpe, M., and Lorito, S.: Analytical Tsunami Fragility Modelling for Building Classes Using Tsunami Time-History Simulations, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11453, https://doi.org/10.5194/egusphere-egu25-11453, 2025.
Global Tsunami Model (GTM) Probabilistic Tsunami Hazard Assessment (PTHA) is one of the Pilot Demonstrators (PD) of the EuroHPC JU ChEESE-2P project, within the scope of GTM organization. As an updated version of Davies et al. (2018) global model, this new one will include enhanced features such as stochastic slip models, spatially higher resolution of the calculation points with particular attention to relatively small islands, and the contribution of tides and long-term sea level variations, among other things. We also aim to make it interoperable with the GEM OpenQuake tools and consistent with similar seismic hazard models.
As an initial step, a tsunami Green's functions (GF) database for subduction zones (meshed as quadrilateral subfaults) in the Pacific Ocean was created on to CINECA Leonardo supercomputer. This database is being used to set up simulations of tsunami GFs on a global grid, using sources in the Pacific Ocean, which will be used for the GTM PTHA. An optimal trade-off between the available computational and storage resources and the resolution/duration and accuracy of the numerical simulations is being sought for. We are also using real events’ tsunami records to determine whether the initial model settings are adequate for accurately modelling observed data, following the approach by Davies (2019).
We will also report about the testing the new version of Tsunami-HySEA that implements the computation of initial conditions from triangular subfaults (Nikkhoo & Walters, 2015), including the contribution of the horizontal deformation (Tanioka & Satake, 1996), and the “Nosov” filter (Abbate et al., 2024).
Davies, G., Griffin, J., Løvholt, F., Glimsdal, S., Harbitz, C., Thio, H. K., et al. (2018). A global probabilistic tsunami hazard assessment from earthquake sources. Geological Society, London, Special Publications 456, 219–244. doi: 10.1144/sp456.5
Davies, G. (2019). Tsunami variability from uncalibrated stochastic earthquake models: tests against deep ocean observations 2006–2016. Geophysical Journal International, 218(3), 1939-1960.
Nikkhoo, M., & Walter, T. R. (2015). Triangular dislocation: an analytical, artefact-free solution. Geophysical Journal International, 201(2), 1119-1141.
Tanioka, Y., & Satake, K. (1996). Tsunami generation by horizontal displacement of ocean bottom. Geophysical research letters, 23(8), 861-864.
Abbate, A., González Vida, J. M., Castro Díaz, M. J., Romano, F., Bayraktar, H. B., Babeyko, A., & Lorito, S. (2024). Modelling tsunami initial conditions due to rapid coseismic seafloor displacement: efficient numerical integration and a tool to build unit source databases. Natural Hazards and Earth System Sciences, 24(8), 2773-2791.
How to cite: Bayraktar, H. B. and the GTMPTHA Working Group: Testing the simulation setup for the GTM PTHA, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19584, https://doi.org/10.5194/egusphere-egu25-19584, 2025.
This study investigates the complex dynamics of tsunami wave run-up, emphasizing nonlinear wave behavior. Utilizing the self-manipulated interFoam solver, we analyze various factors that affect run-up, including surf-similarity, wave non-linearity and beach slope. Our findings reveal a consistent pattern in the variation of run-up height with surf-similarity across different levels of wave non-linearity: an initial increase followed by a decrease as surf-similarity intensifies. Waves with low surf-similarity tend to exhibit significant run-up accompanied by wave breaking, while those with high surf-similarity demonstrate gentler and more prolonged run-up and run-down processes. Under constant surf-similarity conditions, tsunamis on mild slopes break more readily, resulting in lower run-up heights compared to those on steep slopes. Additionally, waves characterized by higher non-linearity are more likely to break than those with similar surf-similarity but lower non-linearity.
We calibrate the analytical solution proposed by Madsen & Schäffer (2010) and introduce semi-empirical methods for the conservative estimation of run-up height and velocity, along with an empirical formula for estimating swash periods. This methodology leverages wave data collected along a sloping beach, thereby eliminating the need for arbitrary inputs from the beach’s toe or offshore regions. Importantly, our methods demonstrate effectiveness in estimating run-up heights for waves with non-linearity up to 1.3, indicating their applicability across a broad spectrum of conditions. Despite certain limitations, the proposed methods and formulas represent valuable contributions to tsunami forecasting and hazard assessment, offering insights and alternative pathways for further research in this complex field.
How to cite: Xie, P.: CFD modeling of nonlinear tsunami wave run-up dynamics, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17206, https://doi.org/10.5194/egusphere-egu25-17206, 2025.
Destructive tsunamis are often triggered by shallow coseismic ruptures in subduction zones, making the rapid determination of rupture depth crucial for issuing timely tsunami warnings and mitigating associated hazards. To address this challenge, we propose a deep learning framework for the rapid classification of rupture depth (shallow or deep) based on high-rate GNSS data.
Using the Alaska subduction zone as a case study, we generated nearly 10,000 synthetic earthquake scenarios to overcome the scarcity of real-world megathrust earthquake records. From these simulations, we constructed a comprehensive near-field GNSS three-component displacement waveform database. Leveraging this dataset, we designed a deep learning neural network that extracts critical seismic signal features from high-rate GNSS data to accurately classify rupture depth. The model achieved over 90% accuracy, precision, and recall on the test set.
We applied the model to the 2021 Mw 8.2 Alaska earthquake and successfully identified it as a deep rupture, with a processing time of approximately 20 ms. Additionally, through transfer learning, we extended the model to the Sumatra subduction zone and successfully identified the 2010 Mw 7.8 Mentawai earthquake as a shallow rupture. This study provides a valuable reference for enhancing the reliability of tsunami early warning systems.
How to cite: Cui, W., Chen, K., and Zhang, N.: Rapid Identification of Rupture Depth in Subduction Zone Earthquakes Based on High-Rate GNSS Using Deep Learning, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8246, https://doi.org/10.5194/egusphere-egu25-8246, 2025.
The complexity of coseismic slip distributions plays a pivotal role in shaping tsunami hazards from both near and distant sources. Recent research underscores the significance of large shallow slips in tsunamigenic earthquakes, driven by dynamic amplification near the free surface and variable frictional conditions. Several novel methods are being proposed to incorporate depth-dependent features and shallow slip amplification in subduction earthquake models, possibly ensuring balanced long-term total slip across seismic cycles. This allows to incorporation of these slip models in Probabilistic Tsunami Hazard Assessment (PTHA). Applying this approach to the central and eastern Mediterranean using 3D subduction geometries, their findings revealed increased probabilities for larger tsunami inundation heights, underscoring the need for improved hazard assessments in global subduction zones.
Depth-dependent rigidity variations also critically influence initial tsunami size estimates, highlighting the necessity of consistent rigidity models for accurate tsunami hazard analysis. Expanding previous models, our research incorporates both depth-dependent rigidity and stress drop into tsunami hazard modelling. This refinement aligns with common observations that shallow subduction earthquakes exhibit longer source durations than deeper events. By addressing inconsistencies arising from varying only rigidity, our enhanced methodology offers tsunami hazard curves grounded in a more physically robust seismic source model.
Our study emphasizes the role of stress drop variability across three defined rigidity gradients with depth, ranging between the constant stress drop end-member model of Bilek & Lay (1999) and the Preliminary Reference Earth Model (PREM). We apply this approach to the Calabrian, Hellenic, and Cyprus subduction zones in the Mediterranean. Given a fixed seismic moment, rupture duration is influenced by rupture size and propagation velocity, in turn, related to the stress drop and rigidity, respectively. By adjusting rupture length and width along the dip, we calibrate our model to observed rupture durations, capturing the stress drop variation with depth. Differently from models imposing fixed stress drop values, ours prioritizes calibrating this gradient to achieve a physically more consistent representation of earthquake sources.
This study explores the extent to which detailed modelling of stress drop variability, shallow slip amplification, and depth-dependent rigidity affect tsunami hazard curves within the Mediterranean basin, with a particular focus on the probabilities of larger inundation heights. The results contribute to refining earthquake source modelling for tsunami forecasting, benefiting both PTHA and early warning systems like Probabilistic Tsunami Forecasting. Parallelly, we are also testing the consistency of this model with tsunami observations of past Pacific events.
How to cite: Vishnu, K. N., Scala, A., Lorito, S., Romano, F., Tonini, R., Volpe, M., Bayraktar, H. B., and Festa, G.: Depth-dependent stochastic slip models governed by stress drop and rigidity variations in subduction zones: advancements in probabilistic tsunami hazard analysis. , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12538, https://doi.org/10.5194/egusphere-egu25-12538, 2025.
Probabilistic workflows are indispensable for assessing the overland tsunami hazard and risk, due to the infrequency and limited historical observations of tsunamis. However, these workflows are computationally demanding because they require a large number of simulations to capture uncertainty of the phenomena. This study leverages machine learning (ML) emulators to address this challenge by directly predicting hazard and risk metrics, bypassing the need for extensive numerical simulations for the inundation phase.
The ML emulators are trained to predict high-resolution hazard metrics onshore (e.g., maximum inundation depth) and risk metrics (e.g., expected damage or loss) using offshore waveforms and local deformation fields as inputs. A database of tsunamigenic earthquakes in the Mediterranean Sea, reflecting substantial variability in source mechanisms and locations, was used for training and validation. For a test site in Sicily, Italy, the emulator demonstrated robust performance with a training set of ~1,600 events, achieving a 30-fold reduction in computational cost compared to traditional probabilistic tsunami hazard assessment (PTHA) workflows.
In the aftermath of tsunami event, such ML emulators can be used to directly provide rapid estimates on the expected damage and losses at different disaggregation, while evaluating many different scenarios due to the uncertainty in the characterization of the earthquake source in the early stages of after the earthquake event.
How to cite: Ragu Ramalingam, N.: Machine Learning Approaches for Tsunami Hazard and Risk Assessment , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19939, https://doi.org/10.5194/egusphere-egu25-19939, 2025.
Site-specific Probabilistic Tsunami Hazard Assessment (PTHA) is a powerful tool for coastal planning against tsunami risk. However, its typically high computational demands led to the introduction of a Monte Carlo Stratified Importance Sampling (SIS) approach, which selects a representative subset of scenarios for numerical inundation simulations. We here empirically validate this sampling approach, for the first time to our knowledge, using an existing extensive dataset of numerical inundation simulations for two coastal sites in the Mediterranean Sea (Catania and Siracusa, both located in Sicily, Italy). Moreover, we propose a modified importance sampling function to prioritise seismic tsunami scenarios based on their arrival time at an offshore point near the target site, in addition to their wave amplitude and occurrence rate as leveraged in the previous work. This sampling function is applied separately in each earthquake magnitude bin, and allows denser sampling of near-field earthquakes to whose variations tsunamis are very sensitive.
We compare the confidence intervals of the offshore PTHA estimates obtained with the new and the original importance sampling functions. Then, we benchmark our onshore PTHA estimates obtained with both functions against the inundation PTHA calculated using the full set of scenarios. We also test the assumption that onshore random errors follow a normal distribution, as found previously for the offshore case. As a result of the benchmarks, we find that the SIS approach works satisfactorily. Introducing the arrival time as an additional sampling factor enhances the precision of the estimates of both the mean and the percentiles for the two coastal sites considered. With this modification it is possible to deal efficiently with heterogeneous near-field earthquake sources involving coastal deformation at Catania and Siracusa, in addition to regional crustal and subduction sources. By comparing the sampling errors with the model (epistemic) uncertainty, an optimal trade-off between the number of simulations employed and the uncertainty of the PTHA model can be found, even for such a complex situation. A relatively small number of scenarios, on the order of a few thousand, is sufficient to perform site-specific PTHA for practical applications. These numbers correspond to 4-8\% of the already reduced ensembles used in previous assessments at the same sites.
How to cite: Abbate, A., Davies, G., Lorito, S., Kalligeris, N., Romano, F., Tonini, R., and Volpe, M.: Importance sampling of seismic tsunami sources with near-field emphasis for inundation PTHA: benchmarking with complete ensembles, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19787, https://doi.org/10.5194/egusphere-egu25-19787, 2025.
The Global Tsunami Model (GTM) global-scale Probabilistic Tsunami Hazard Assessment (PTHA) is one of the Pilot Demonstrators (PD) of the EU ChEESE-2P project, which would represent an update of the previous global tsunami hazard model proposed by Davies et al. (2018). Since it is a PTHA for earthquake-generated tsunamis, it is important that its input seismic model is consistent with the one used for Probabilistic Seismic Hazard Analysis (PSHA) at comparable scales and affecting the same locations.
The GTM and the Global Earthquake Model (GEM) organizations then started collaborating to improve the interoperability of the tools used for PTHA and PSHA, and of the input and output data and models. This could benefit the end-users since both the shaking and the inundation result from the same causative phenomenon - the earthquake in this case.
Moreover, the GEM OpenQuake (OQ) engine for seismic hazard and risk assessment provides an opportunity to compare the GTM tools with a well-tested software platform that uses accepted standards.
In this contribution we present the first results of a sensitivity analysis of the PTHA results to the use of different earthquake occurrence models for the same seismogenic source zone, and to the use of different tools and codes for the generation of earthquake rupture catalogues, for the tsunami propagation, and for the aggregation of the hazard results. To this end, we use different combinations of the data, tools and codes from those of Davies et al. (2018) and the Australian PTHA (Davies, 2019), the GTM ones (e.g. Gibbons et al., 2020), and the OQ ones (Pagani et al., 2014).
Davies G., et al., 2018. A global probabilistic tsunami hazard assessment from earthquake sources. Geological Society, London, Special Publications 456, 219–244. doi: 10.1144/sp456.5
Davies G., 2019. "A new probabilistic tsunami hazard assessment for Australia." Australasian Coasts and Ports 2019 Conference: Future directions from 40 S and beyond, Hobart, 10-13 September 2019.
Gibbons S.J., et al., 2020. “Probabilistic Tsunami Hazard Analysis: High Performance Computing for Massive Scale Inundation Simulations”. Front. Earth Sci. 8:591549. doi: 10.3389/feart.2020.591549
Pagani M., et al., 2014. OpenQuake Engine: An open hazard (and risk) software for the Global Earthquake Model, Seismol. Res. Lett., 85, 3, 692-702, doi:10.1785/0220130087.
How to cite: Cascone, V., Bayraktar, B., Basili, R., Crowley, H., Gibbons, S., Johnson, K., Lorito, S., Løvholt, F., Pagani, M., Romano, F., Tonini, R., and Volpe, M.: The GTM global PTHA: towards interoperability with the GEM OpenQuake engine, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19264, https://doi.org/10.5194/egusphere-egu25-19264, 2025.
On 30 October 2020, the island of Samos (Greece) and the region of Izmir (Türkiye) were hit by a powerful earthquake followed by a tsunami that spread across the Aegean Sea. The magnitude M7.0 earthquake caused severe damage and more than 100 deaths in both countries, including one death in Türkiye due to the tsunami. This disaster was yet another reminder of the Mediterranean region's vulnerability to seismic and tsunami hazards. It highlighted the critical significance of prevention and preparedness in mitigating the impacts of natural hazards.
Following this disaster, the first Greek community was recognised as Tsunami Ready alongside the efforts of the Intergovernmental Oceanographic Commission of UNESCO (UNESCO-IOC) to reinforce the resilience of coastal communities in the Northeast Atlantic, Mediterranean, and Connected Seas (NEAM) region and around the globe. The Tsunami Ready recognition of the town of Samos was achieved through the CoastWAVE Project, coordinated by UNESCO-IOC and funded by DG-ECHO of the European Commission to enhance the resilience of NEAM coastal communities to tsunamis and other sea-related hazards. Aligned with the goals that UNESCO-IOC has set through the Ocean Decade Program, the project focused on piloting the Tsunami Ready Recognition Program (TRRP) standards and guidelines. As a result of the project, selected communities, including the town of Samos, gained more awareness of tsunami risks, improved tsunami risk governance, and were recognized as Tsunami Ready.
In the case of Greece, key project tasks involved tsunami awareness activities, establishing a National Tsunami Ready Board, hazard and evacuation mapping, developing local protocols and Standard Operational Procedures, and testing them through a local exercise involving stakeholders at local, national, and international levels. The strong collaboration between tsunami experts, local authorities, emergency management agencies, and other stakeholders, coordinated by the National Observatory of Athens and the Municipality of Eastern Samos, proved to be the cornerstone of success. The strategic alliance between science and emergency management brought about more comprehensive preparedness and risk reduction efforts and underlined the importance of knowledge-sharing.
The town of Samos managed to fulfill all the UNESCO-IOC TRRP criteria to become Tsunami Ready, however, this was only the first step. To maintain the resilience gained in the process of becoming Tsunami Ready, Samos needs to keep fulfilling the indicators of the TRRP to remain prepared and ready to respond against the threat of tsunamis. To this end, the continuous commitment of local and national stakeholders is essential in building a more resilient future against tsunamis and other natural hazards in the NEAM region. Education, preparedness, and community involvement are some of the key elements that can be exploited to enhance safety and reduce risks associated with tsunami events.
We will present the approaches followed and the activities undertaken to fulfill the Tsunami-Ready indicators in the town of Samos, along with the specificities and challenges faced in this first TRRP implementation in Greece.
How to cite: Kalligeris, N., Daskalaki, E., Dandoulaki, M., Plessa, A., Papageorgiou, A., Melis, N. S., Lentas, K., Skanavis, V., Ktenidou, O.-J., Gkika, F., and Charalampakis, M.: The first Tsunami Ready community in Greece: Samos town (Samos island, Northern Aegean Sea), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15219, https://doi.org/10.5194/egusphere-egu25-15219, 2025.
On 6 February 2023 the region between southern Turkey and northern Syria was hit by a devastating earthquake sequence, starting with a Mw = 7.8 event at 01:17:34 UTC on the Eastern Anatolian Fault (EAF), followed by a Mw = 7.5 at 10:24:29 UTC along the Sürgü Fault (SF). Due to the comparable size of the two events and the mutual positions (on two separate structures, EAF and SF), they are considered “doublet” earthquakes. Aftershocks occurred for few weeks after the first mainshock (Mw=7.8) and many different coseismic and secondary effects accompanied the seismic sequence. The Mw=7.8 event was followed by a modest tsunami that was observed on few coastal Tide Gauges (TGs) in the eastern Mediterranean. Historical tsunami catalogues contain very few entries of past tsunamis in this area of the Levantine coast. In this work, we aim to constrain the nature and location of the tsunami source through numerical simulations. Two generation mechanisms are considered: the first involves the activation of an offshore tectonic source, while the second considers submarine landslides. The latter are modelled using a combination of two Gaussian functions with opposite polarity as the analytical initial condition. Several scenarios, based on both tectonic and mass movement sources, are tested employing JAGURS, a numerical code that computes tsunami propagation and inundation on the basis of the long wave approximation. The results of the simulations are compared with the observations available, provided by the tide gauge stations of Gazimagusa/Famagosta (Cyprus), Arsuz (Turkey), Erdemli (Turkey) and Tasucu (Turkey), allowing for the identification of a source area capable of reproducing the main characteristics of the observed TG records during the first minutes following the tsunami's arrival. Whatever the type of source considered, none of the tested scenarios is able to reproduce all the main observed characteristics (arrival time, period, polarity and amplitude of the first peak) of the recorded waveforms. At this stage, we favour the hypothesis of a complex generating mechanism, combining a predominant role played by one or more submarine landslides, possibly “tuned” by a contribution from coseismic offshore ruptures.
How to cite: Saturnino, C., Angeli, C., Zanetti, M., Zaniboni, F., and Armigliato, A.: The 6th February 2023 tsunami in the Eastern Mediterranean: on the origin of the event, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17435, https://doi.org/10.5194/egusphere-egu25-17435, 2025.
This study investigates tsunami generation induced by landslides on the island of Ischia, located in the Tyrrhenian Sea, an area particularly vulnerable due to its unstable slopes and proximity to densely populated coastlines. The research utilizes a dataset of 165 documented landslide events, which were analysed to reconstruct "ideal" landslides based on parameters such as area, volume (ranging from 10³ m³ to over 10⁷ m³), average length and width, centroid coordinates, and deposit thickness. These parameters were used to create representative geometric models for numerical simulations with the COMCOT model. The COMCOT model, known for its ability to simulate tsunami generation, propagation, and coastal interaction, was applied to landslides of varying sizes and orientations. Eleven observation points along the island's coast were defined to track changes in wave height, energy distribution, and the temporal evolution of impacts. Results show that large-volume landslides generate significantly higher and more destructive waves, with local amplifications occurring in areas of irregular bathymetry, such as coves and bays. Simulations revealed that landslides oriented to the north produced waves reflected towards the open sea, reducing direct impact on the coast. In contrast, landslides oriented to the east generated higher waves with direct propagation towards the Gulf of Naples, increasing the risk of flooding in ports and urban areas. The largest waves, exceeding 10 meters in height, were observed in scenarios involving large-volume landslides, underscoring the destructive potential of such events. This study introduces a novel methodological approach by modeling landslides based on real data from a large database of different landslides in the area.
How to cite: Caporale, G., Grezio, A., and Selva, J.: Tsunami Induced by Landslides on the Island of Ischia: numerical simulations and hazard analysis, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11633, https://doi.org/10.5194/egusphere-egu25-11633, 2025.
Tsunami propagation in an ice-covered sea
Yakovlev V.V., Tkachenko V.O., Bondar V.V., Goncharenko T.B.
Institute of Hydromechanics, of National Academy of Sciences of Ukraine
When a tsunami wave propagates into a sea area covered with solid ice, the part of the wave that has passed under the ice cover will be affected by the elastic properties of the ice sheet. These properties radically change the nature of the tsunami wave propagation.
A long-wave nonlinear dispersion model describing the propagation of tsunami flexural-gravity waves in a continuous ice sheet floating on the sea surface is constructed by expanding the initial three-dimensional problem of hydroelastic oscillations of the system "elastic plate – layer of ideal incompressible fluid of variable depth" in a small parameter. The model takes into account the effects of nonlinear dispersion of fluid, as well as inertia, elasticity and geometrically nonlinear deflection of the plate. Based on the obtained equations, a hierarchical sequence of simpler models is constructed, generalizing the equations of Peregrine, Boussinesque and Korteweg-de Vries, known from the theory of surface waves, to the case of flexural-gravity waves. In the particular case of the generalized Korteweg-de Vries equation, which describes the propagation of tsunami waves, exact solutions are constructed and analyzed, describing the propagation of solitons and cnoidal waves in the sea covered with solid ice. It is shown that flexural-gravity tsunami waves have some mirror properties compared to tsunami waves on water. In relation to the soliton, this means that without changing the shape, a depression, not a hump, as in the case of a tsunami wave on water, propagates, and the speed of its propagation decreases with increasing amplitude, not increases. In addition, the characteristics of flexural-gravity tsunami waves are determined by the amplitude and dispersion of the flexural rigidity of the plate and do not depend on the dispersion of water and the inertial properties of the ice cover.
For the generalized Korteweg-de Vries equation, to which the model of tsunami wave propagation in the sea covered with solid ice is reduced, the regions of variation of the physical parameters of the problem are identified, where different types of soliton-like solutions of this equation can exist. The nature of the eigenvalues for different ratios of the physical parameters of the problem is investigated and the region of variation of the parameters is determined, in which stationary solutions of the classical solitary wave type can take place.
How to cite: Yakovlev, V., Tkachenko, V., Bondar, V., and Goncharenko, T.: Tsunami propagation in an ice-covered sea , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-491, https://doi.org/10.5194/egusphere-egu25-491, 2025.
Meteotsunamis, tsunami-like waves triggered by rapid atmospheric pressure disturbances, can result in significant coastal damages. This study introduces a machine learning (ML) framework for predicting meteotsunami occurrences along the Portuguese coast, using both atmospheric pressure records and tide gauge data collected from 2010 to 2020. A methodology is proposed to construct a structured dataset of inputs and targets from continuous meteorological and sea-level observations, yielding an imbalanced dataset with a meteotsunami-to-nonevent ratio of approximately 1:60. To address this imbalance, class weighting and an ensemble strategy aggregating predictions across multiple observatories were implemented in the ML framework.
The prediction model employs an encoder-decoder architecture, integrating Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) layers. Results demonstrate the model's effectiveness in capturing the complex dynamics of meteotsunami formation and propagation with accuracy and reliability for operational forecasting. Future research will focus on incorporating additional meteorological variables such as wind speed and direction, expanding the spatial and temporal coverage of data, and further refining prediction capabilities to enhance meteotsunami early warning systems and mitigate meteotsunami-related risks.
How to cite: Kim, J. and Omira, R.: Machine Learning for Meteotsunami Prediction: A Case Study on the Portuguese Coast, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8785, https://doi.org/10.5194/egusphere-egu25-8785, 2025.
An analysis of the causes of the occurrence of meteotsunamis, which were registered on May 7, 2007 on the northern part of the Bulgarian coast, on June 14, 2014 in the Odessa region and in the Ilyichevsk (Chernomorsk) port (Sukhoi Estuary), on July 19, 2017 in the waters of the Belosarayskaya Spit in Azov, showed that all these phenomena occurred under similar macroscale synoptic conditions situations over south-eastern Europe.
At the same time, the required morphometric conditions were present at the points of meteotsunami recording: a low rate of depth decrease towards the coast, the location and structure of the coast in all three cases suggested the possibility of a long sea wave arriving from the open sea from a distance of 130-200 km and, accordingly, the occurrence of Proudman resonance. In all cases, the local nature of the phenomenon was noted: the length of the wave crest along the front did not exceed several tens of kilometers.
Analysis of the maps of the high-altitude geopotential (850 and 500 hPa) for the Azov-Black Sea region on the indicated dates shows a classic picture of the frontal interaction between the Asia Minor Depression (with dry and very warm air of African origin) and the cold and humid (polar) air of the anticyclone over Eastern Europe. Such fronts are a source of atmospheric instability and wind strengthening at all levels.
A comparative analysis of synoptic maps for the above dates showed a fairly good qualitative correspondence between the structure of surface pressure fields and the location of fronts; satellite information also showed the presence of zones of powerful convective cloudiness over the Azov-Black Sea region. An important feature of the synoptic situation is the instability line over the western or central parts of the Black Sea, which indicates the presence of a ridge of cumulonimbus clouds (Cb) and the existence of powerful convective movements that can reach the stratosphere - overshotting helps replenish the energy of jet streams. This structure and state of the atmosphere was defined by A. Rabinoviches and J. Šepić with the general term – a "tumultuous atmosphere". Thunderstorm phenomena characteristic of Cb are capable of generating a wide range of internal gravity waves with characteristic periods from 3 to 60 minutes.
The analyzed synoptic conditions during the tsunami were completely favorable for the occurrence and propagation of a possible moving atmospheric gravity disturbance. In all three cases, the Froude number was close to unity, indicating that the conditions for the Proudman resonance were met.
Thus, the combination of synoptic and geographical factors indicates a significant probability of this phenomenon occurring only in certain areas of the Azov-Black Sea region: the western shelf of the northwestern part of the Black Sea, including relatively deep-water estuaries, and the Belosarayskaya Spit area of the Azov Sea.
How to cite: Matygin, A. and Iakovleva, N.: Synoptic conditions for the generation of meteotsunamis on the shelf of the northwestern Black Sea region and in the Azov Sea, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-450, https://doi.org/10.5194/egusphere-egu25-450, 2025.
Posters virtual: Wed, 30 Apr, 14:00–15:45 | vPoster spot 3
EGU25-20966 | Posters virtual | VPS13
Cataloging historical tsunami marigrams from microfilm imagesWed, 30 Apr, 14:00–15:45 (CEST) | vP3.4
The U.S. NOAA National Centers for Environmental Information (NCEI) has more than 3,700 tsunami marigram (tide gauge) records in both image and paper format, capturing worldwide observations of more than 390 tsunami events from 1854 to 1994. The majority of these tsunami marigram records were scanned to high-resolution digital TIFF images during the U.S. NOAA Climate Data Modernization Program (CDMP) which ran from 2000 to 2011. Additional, uncatalogued physical records exist on microfilm rolls and paper at the David Skaggs Research Center (DSRC) in Boulder, Colorado, USA. For many tsunami events prior to 1994, data resides only on the marigram records, making them of great historical significance. Six of the 13 uncatalogued microfilm rolls have been scanned by NCEI to produce 3,548 TIFF images. During 2025, we will be working to catalog, archive, and make these images discoverable and accessible online. We will identify any duplicates by comparing to the existing catalog of marigrams already archived at NCEI. Given the large number of uncatalogued images, we are exploring automated approaches to harvesting metadata from the images to aid in cataloging. We will present the project background, goals, and initial results of this effort.
How to cite: Sweeney, A. and Radio, E.: Cataloging historical tsunami marigrams from microfilm images, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20966, https://doi.org/10.5194/egusphere-egu25-20966, 2025.
