Enter Zoom Meeting

NH5.1

EDI
Tsunamis: from source processes to coastal hazard and warning

Tsunamis can produce catastrophic damage on vulnerable coastlines, essentially following major earthquakes, landslides, extreme volcanic activity or atmospheric disturbances. After the disastrous tsunamis in 2004 and 2011, tsunami science has been continuously growing and expanding its scope to new fields of research in various domains, and also to regions where the tsunami hazard was previously underestimated.

The spectrum of topics addressed by tsunami science nowadays ranges from the “classical” themes, such as analytical and numerical modelling of different generation mechanisms (ranging from large subduction earthquakes to local earthquakes generated in tectonically complex environments, from subaerial/submarine landslides to volcanic eruptions and atmospheric disturbances), propagation and run-up, hazard-vulnerability-risk assessment, especially with probabilistic approaches able to quantify uncertainties, early warning and monitoring, to more “applied” themes such as the societal and economic impact of moderate-to-large events on coastal local and nation-wide communities, as well as the present and future challenges connected to the global climate change.

This session, co-organized with OS4, SM4, GMPV9, GM and AS, welcomes multidisciplinary as well as focused contributions covering any of the aspects mentioned above, encompassing field data, geophysical models, regional and local hazard-vulnerability-risk studies, observation databases, numerical and experimental modeling, real time networks, operational tools and procedures towards a most efficient warning, with the general scope of improving our understanding of the tsunami phenomenon, per se and in the context of the global change, and our capacity to build safer and more resilient communities.

Co-organized by GM6/OS4/SM4
Convener: Alberto Armigliato | Co-conveners: Ira Didenkulova, Hélène Hébert, Lyuba DimovaECSECS
Presentations
| Wed, 25 May, 08:30–11:44 (CEST), 13:20–14:23 (CEST)
 
Room 1.61/62

Wed, 25 May, 08:30–10:00

Chairpersons: Alberto Armigliato, Ira Didenkulova, Joern Behrens

08:30–08:37
|
EGU22-39
|
ECS
|
|
On-site presentation
Ramtin Sabeti and Mohammad Heidarzadeh

Subaerial landslide-generated waves are among natural hazards that have attracted attention in recent years, in particular after the 2018 Anak Krakatau volcanic tsunami (Indonesia), which left a death toll of over 450. This has increased the application of physical modelling on subaerial landslide tsunamis to cope with the risks of such hazards and to develop knowledge of their generation mechanisms. Physical experiments in two-dimensional flumes are generally more cost-efficient, less time consuming and allow better control on the set-up. As a result, landslide–tsunamis are considerably investigated in 2D rather than in 3D. However, it is important to note that 2D physical modelling of subaerial landslide–tsunamis could be associated with some uncertainties and may slightly overestimate the wave amplitudes. By using 3D physical models, it is possible to investigate wave amplitude attenuations in both radial and angular directions, which would improve the understanding of wave propagation. In this research, we conduct 2D and 3D experiments on subaerial landslide tsunamis. The physical experiments were conducted in a 2.5 m wide, 0.50 m deep and 2.5 m long wave basin at the Brunel University London (UK). The experimental setup included five different slope angles (i.e. 25o,35o,45o,55o and 65o). The solid blocks had four different volumes in a range of 0.5×10-12 km3-3.0×10-12 km3. The generated water waves were measured using six precision capacitance wave gauges located in both near- and far-fields. The 2D and 3D results are compared to quantify the effects of dimensions on the wave amplitudes and attenuations.

How to cite: Sabeti, R. and Heidarzadeh, M.: Three-dimensional physical modelling of subaerial landslide-generated waves and comparison with two-dimensional experiments, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-39, https://doi.org/10.5194/egusphere-egu22-39, 2022.

08:37–08:44
|
EGU22-3912
|
Virtual presentation
Pablo Poulain et al.

We investigate the dynamics and deposits of granular flows and the amplitude of the generated water waves using the depth-averaged shallow numerical model HySEA, both at the lab- and field scales. We investigate the different sources of errors by quantitatively comparing the simulations with (i) six new laboratory experiments of granular collapses in different conditions (dry, immersed, dry flow entering water) and slope angles, and (ii) numerical simulations made with the code SHALTOP that describes topography effects better than most landslide-tsunami models. In the laboratory configurations, at the limit of the shallow-approximation in such models, we show that topography and non-hydrostatic effects are crucial. However, when empirically accounting for topography effects by artificially increasing the friction coefficient and performing non-hydrostatic simulations, the model is able to reproduce the granular mass deposit and the waves recorded at gauges located at a distance of more than 2-3 times the characteristic dimension of the slide, with an error ranging from 1 % to 25 % depending on the scenario, without any further calibration. Taking into account this error estimation, we simulate landslides that occurred on Montagne Pelée volcano, Martinique, Petites Antilles as well as the generated waves. Results support the hypothesis that large flank collapse events in Montagne Pelée likely occurred in several successive sub-events. This result has a strong impact on the amplitude of the generated waves, and thus on the associated hazards. In the context of the on-going seismic volcanic unrest at Montagne Pelée volcano, we calculate the debris avalanche and associated tsunami for two potential flank-collapse scenarios.

How to cite: Poulain, P., Le Friant, A., Mangeney, A., Viroulet, S., Fernandez-Nieto, E., Castro Diaz, M., Peruzzetto, M., Grandjean, G., Bouchut, F., Pedreros, R., and Komorowski, J.-C.: Performance and limits of a shallow model for landslide generated tsunamis: from lab experiments to simulations of flank collapses at La Montagne Pelée (Martinique), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3912, https://doi.org/10.5194/egusphere-egu22-3912, 2022.

08:44–08:51
|
EGU22-4804
|
ECS
|
Virtual presentation
Matthias Rauter et al.

Subaerial landslides are among the most complex sources for tsunamis, as several complex processes occur simultaneously in various regimes, with multiple phases interacting. The simulation and prediction of these events is respectively difficult.

We will present a three-dimensional multiphase model (granules, air, water) that considers the  effects and properties that we deem most important: (i) a sharp water-air interface with low diffusivity, (ii) granular rheology for the landslide, (iii) differentiation between effective pressure and pore pressure, as well as (iv) porosity, dilatancy and permeability. No depth-integration or other form of simplification is applied. The resulting mathematical model is solved with the fluid dynamics toolkit OpenFOAM.


Many effects and processes that are lost in depth-integrated models are directly simulated in our approach. This allows the simulation of complex events with a relatively simple model, however for a large computational cost. The model parameters are widely intrinsic material parameters, which promises a prediction of events without significant parameter optimizations.

We will show results for small scale experiments as well as for a well documented real scale event and will give an outlook on further developments and remaining problems.

How to cite: Rauter, M., Viroulet, S., Gylfadóttir, S. S., Løvholt, F., and Fellin, W.: Granular porous landslide tsunami modelling with OpenFOAM, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4804, https://doi.org/10.5194/egusphere-egu22-4804, 2022.

08:51–08:58
|
EGU22-8236
|
On-site presentation
Glauco Gallotti et al.

Fjord environments are subject to submarine mass wasting events due to their steep slopes, high sedimentation rates, and tectonic activity driven by glacial-isostatic rebound. In specific cases, these events can generate tsunami waves whose coastal heights are strongly influenced by the physiography, both subaerial and submarine, of the fjord. Here we present modeling simulations of a potential tsunami initiated by a submarine landslide in Pangnirtung Fjord, eastern Baffin Island (Nunavut, Canada). Pangnirtung Fjord, a 43 km long, 1 to 3 km wide, and 165 m deep fjord, is fed by numerous rivers that transport sediment from the surrounding high-relief, partially glaciated landscape. Collapse of the Kolik River delta, situated directly across from the hamlet of Pangnirtung, is the likely cause of the largest submarine landslide (2.1 km2) identified in the fjord using multibeam bathymetric data and 3.5 kHz sub-bottom profiles collected in 2019. The mapped landslide extends across the flat basin and features a blocky deposit directly downslope of the delta. The landslide dynamics, the consequent water waves generation and propagation were simulated by means of codes developed by the Tsunami Research Team of Bologna University. The landslide parameters characterizing the downslope motion have been retrieved by matching the landslide dynamics with the observed deposit. As the landslide impulses to the water column are considered, the propagation of the waves inside the fjord is determined through the shallow water approximation of the Navier-Stokes set of equations. The waves reach the hamlet (3.5 km from the landslide source) in 200 s, and the surrounding fjord coasts in approximately 800 s. Maximum wave height values of approximately 2 m were modeled and used to construct an inundation map for the area, over a 2 m regularly spaced grid for the hamlet of Pangnirtung.

How to cite: Gallotti, G., Sedore, P., Armigliato, A., Normandeau, A., Maselli, V., and Zaniboni, F.: Submarine landslide tsunamis in fjord environments: the case of Pangnirtung Fjord, eastern Baffin Island (Nunavut, Canada), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8236, https://doi.org/10.5194/egusphere-egu22-8236, 2022.

08:58–09:05
|
EGU22-12471
|
Virtual presentation
Enrique Fernandez-Nieto et al.

Depth-averaged models, such as the Savage-Hutter model with Coulomb or Pouliquen friction laws, are usually considered to simulate aerial and submarine avalanches. In particular,  submarine avalanches can be the source of a tsunami. These models are presented in local coordinates over the topography or a reference bottom. We show in this work that  classical models do not in some cases preserve the physical threshold of motion. On the one hand, the simulated granular mass can start to flow  even if the slope angle of its free surface is lower than the repose angle of the granular material involved. On the other hand, the granular mass can stay at rest being the slope angle of the free surface higher than the repose angle of the material. Several numerical tests are presented  to illustrate these problems related to classical depth averaged models. In this work we also propose an initial correction which ensures that the model preserves, up to the second order, the physical threshold of motion defined by the repose angle of the material. Several numerical tests are presented, by comparing also with experimental data to illustrate the effect of the proposed correction.

How to cite: Fernandez-Nieto, E., Bouchut, F., Delgado Sánchez, J. M., Narbona-Reina, G., and Mangeney, A.: A bed pressure correction for depth-averaged granular flow models to ensure the physical threshold of motion, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12471, https://doi.org/10.5194/egusphere-egu22-12471, 2022.

09:05–09:12
|
EGU22-8520
Ira Didenkulova et al.

Studies of the influence of coast roughness on run-up height have numerous applications to tsunami problem. It happens when tsunami propagates over the urban area and houses and coastal structures represent roughness elements, which help to dissipate wave energy and reduce maximum tsunami inundation and at the same time can break due to tsunami loading. In this paper we focus on this topic from both points of view and study experimentally and numerically reduction of wave run-up height due to the bed roughness and corresponding wave loading on roughness elements.

Experiments have been performed in a 307 m long, 7 m deep and 5 m wide Large Wave Flume in Hannover, Germany. The experimental setup contained a 251 m long section of the constant depth, which was kept at the depth of h = 3.5 m during all tests, and a 1:6 slope section. A total of 16 wave gauges were mounted along the flume to reconstruct the incident wave field and to study its nonlinear deformation. During the tests, two video cameras and a capacitance probe were used to measure wave run-up on a sloping beach. Two cameras were set up to film the surf zone. One video record was used to calibrate the run-up data measured by the capacitance probe. An additional video record was used to determine the shape of the water surface, which was illuminated by a laser sheet along the direction of wave propagation.

Logs with rectangular 10×10 cm cross-section were used as roughness elements and the force acting on logs was recorded. Two logs were equipped with force transducers; one located at the unperturbed shoreline 272 m and the one located at 276 m mark. Four roughness configurations were considered, with logs every 1 m, 2 m, and 4 m which was compared to the smooth, zero log baseline condition. Waves of different height, period and shape have been used as input signals.

Experimentally shown, that run-up height has a strong non-linear dependence on the amplitude of incident wave and the number of roughness elements. Force acting on the roughness elements is related to the amplitude of the incident wave during the run-up phase and is defined by the flowing down near-slope layer when the bulk of the fluid recedes. At higher wave amplitudes, the average force (total momentum) imposed by roughness elements on the fluid is directed up the slope

Described experiments have been used to validate two numerical models (nondispersive shallow water model and dispersive model based on modified Peregrine equations) and to evaluate the potential of these models to simulate wave attenuation due to sea bed roughness. To model the bottom friction, we used both Manning’s and Chezy’s roughness laws. The results of this work are also discussed.

How to cite: Didenkulova, I., Abdalazeez, A., Dutykh, D., and Denissenko, P.: Effects of coastal roughness on long wave runup, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8520, https://doi.org/10.5194/egusphere-egu22-8520, 2022.

09:12–09:19
|
EGU22-9486
|
ECS
|
|
On-site presentation
Sara Aniko Wirp et al.

The Hellenic Arc is an active seismogenic zone in the Mediterranean Sea that hosted at least two historical M≥8 earthquakes, which both caused destructive tsunamis. The low-angle geometry of its subduction interface could promote shallow slip amplification, enhancing seafloor displacement.
Long-term seismic-probabilistic tsunami hazard assessment (S-PTHA, e.g., Scala et al., 2020) and early warning systems typically rely on kinematic models and Okada's analytical solution to compute static seafloor displacements. The static displacement is then used to source tsunami models. However, the complex interaction of earthquake dynamics and tsunami-genesis may not be fully captured.

We recently demonstrated mechanically consistent dynamic rupture models in generic megathrust settings informed from long-term geodynamic modeling that can provide building blocks toward integrating physics-based dynamic rupture modeling in Probabilistic Tsunami Hazard Analysis (Wirp et al., 21). We here present a range of 3D multi-physics, high-resolution dynamic rupture subduction earthquake scenarios accounting for the complex slab geometry of the Hellenic Arc. We vary hypocenter locations, which leads to a wide range of rupture speeds, extent of shallow fault slip, and moment magnitudes. 
Our dynamic rupture models include highly resolved bathymetry and topography data and detailed knowledge of the tectonic structure of the Hellenic Arc (seismic velocity structure, stresses, and strengths). We use the slab geometry from the European Database of Seismogenic Faults (EDSF, Basili et al., 2013) to create a 3D dynamic rupture scenario that covers great parts of the Mediterranean Sea. The initial conditions in our models are constrained on the subduction zone scale (Ulrich et al., 2021) and specified for the Hellenic Arc region.

Only part of the Hellenic Arc is fully seismically coupled (e.g., Laigle et al., 2004) and most of the convergence is assumed to occur as aseismic creep. We follow Ramos et al. (2021) and apply different friction parameters accounting for high or low coupling of the plate interface.
Our modeling suggests that margin-wide rupture would yield an Mw 9.3 earthquake. More reasonable smaller magnitude earthquakes are obtained by increasing the along-arc complexity of the reference model. Different hypocenter locations result in remarkable differences in shallow fault slip penetrating into velocity-strengthening regions, which translate into strong variations of the final seafloor displacement across scenarios. 
In additional models with partially consolidated and totally unconsolidated sediments (Ulrich et al., 2021) we show that off-fault plastic yielding, which limits shallow fault slip, may drastically increase the seafloor uplift. 
Finally, we explore a novel 3D fully coupled earthquake-tsunami modeling approach (Lotto and Dunham, 2018; Krenz et al., 2021) by adding a water layer to the modeling domain. This enables simulating earthquake dynamics, acoustic waves, and the resulting tsunami simultaneously. The fully coupled model will capture the dynamics of the entire tsunami-genesis in a single simulation, overcoming typical approximations for standard earthquake-tsunami coupling workflows. 

We envision that mechanically consistent dynamic rupture models can provide building blocks toward combined, self-consistent, and physics-based Seismic and Tsunami Hazard Analysis.

How to cite: Wirp, S. A., Ulrich, T., Krenz, L., Bader, M., Lorito, S., and Gabriel, A.-A.: Earthquake scenarios for the Hellenic Arc from 3D dynamic rupture modeling: implications for tsunami hazard, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9486, https://doi.org/10.5194/egusphere-egu22-9486, 2022.

09:19–09:26
|
EGU22-13387
|
ECS
|
|
On-site presentation
Fabian Kutschera et al.

The ~100 km long Húsavík Flatey Fault Zone (HFFZ) in North Iceland is the largest linear transform fault zone in Iceland composed of multiple fault segments that localise both strike-slip and normal movements, agreeing with a transtensional deformation pattern (Garcia and Dhont, 2005). With maximum seismogenic potential larger than Mw 7 and located primarily offshore, the HFFZ subjects several nearby coastal communities to potentially significant tsunami hazard from strong earthquake occurrence on the HFFZ. Namely, tsunami hazard assessment of submarine strike-slip fault systems in transtensional tectonic settings worldwide has received increased attention since the unexpected and devastating local tsunami in the Palu Bay following the 2018 Mw 7.5 Sulawesi earthquake in Indonesia.

Our goal is to carry out a physics-based assessment of the tsunami potential of the HFFZ using both a one-way linked dynamic earthquake rupture and shallow water equations tsunami workflow (Madden et al., 2021) as well as a fully-coupled elastic-acoustic earthquake-tsunami simulation (Krenz et al., SC 2021). We start by simulating physics-based dynamic rupture models with varying hypocenter locations with SeisSol (https://github.com/SeisSol/SeisSol), a scientific open-source software for 3D dynamic earthquake rupture simulation (www.seissol.org). SeisSol, a flagship code of the ChEESE project (https://cheese-coe.eu) and part of the project TEAR (https://www.tear-erc.eu), enables us to explore newly inferred simple and complex fault geometries that have been compiled and proposed in the ChEESE project by using unstructured tetrahedral meshes. The linked workflow uses the time-dependent seafloor displacement output from SeisSol to initialise bathymetry perturbations within sam(oa)²-flash. The dynamically adaptive, parallel software sam(oa)²-flash (https://gitlab.lrz.de/samoa/samoa) solves the hydrostatic shallow water equations (Meister, 2016). Here we consider the contribution of the horizontal ground deformation of realistic bathymetry to the vertical displacement following Tanioka and Satake (1996). Our second approach is based on the recent development of SeisSol which allows us to include a water layer in the earthquake-tsunami simulation to account for fully-coupled 3D elastic, acoustic and tsunami wave generation and propagation simultaneously.


The HFFZ is exposed to a laterally homogeneous regional stress field constrained from seismo-tectonic observations, knowledge of fault fluid pressurisation, and the Mohr-Coulomb theory of frictional failure. We are able to model large Mw 6.7 to 7.3 dynamic rupture scenarios that can generate up to 2m of vertical coseismic offset. Our simulations are controlled by spontaneous fault interaction in terms of dynamic and static stress transfer and rupture jumping across the complex fault network. The models show a dynamic rake rotation of ±20° near the surface, indicating the presence of dip-slip components. Shallow fault slip of up to 8m and off-fault plastic yielding contribute to the tsunami genesis. The sea surface height anomaly (ssha), which is measured at synthetic tide gauge stations along the coastline and defined as the deviation from the mean sea level, provides an estimate about the impact of the tsunami. Our physically informed worst-case tsunami simulation causes a total ssha amplitude of ~1m. We conclude that the HFFZ has the capability to generate localised tsunamigenic earthquakes potentially posing significant hazards to the coastline communities.

How to cite: Kutschera, F., Wirp, S. A., Li, B., Gabriel, A.-A., Halldórsson, B., and Abril, C.: Physics-based earthquake-tsunami modelling of the Húsavík-Flatey transform fault zone in North Iceland, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-13387, https://doi.org/10.5194/egusphere-egu22-13387, 2022.

09:26–09:33
|
EGU22-4840
|
Virtual presentation
Amos Salamon et al.

On July 9th, 551 AD, a destructive earthquake, estimated magnitude 7.5, impacted the Phoenician coast, nowadays Lebanon, Easternmost Mediterranean. Historical accounts describe a sudden withdrawal of the sea from Berytus (Beirut at the time) and other towns along the Phoenician littoral, for a distance of two miles and then return to its normal position, causing many casualties. Critical reading of the historic descriptions raises questions regarding the possible seismogenic and tsunamigenic sources of this catastrophe. Previous researchers presumed inland and offshore seismogenic sources, and submarine earthquake and submarine landslide as tsunami triggers.

Lebanon lies along the Yammouneh restraining bend of the left-lateral Dead Sea Transform (DST), the boundary between the Sinai Sub-Plate (Africa) and Arabia Plate. The bend resulted from a right stepping offset of the DST and thus induces transpressional deformation formed of several thrust faults, such as the recently identified Mount Lebanon thrust (MLT). On the base of extensive geological investigation, marine survey and submarine study (e.g., Elias et al. 2007), the MLT was found to be an active fault that underlies Lebanon and was interpreted to crop out at the seabed, just offshore the coast. It was thus proposed as the source for both the earthquake and the tsunami. Yet, we were puzzled how the significant retreat of the sea and the return to its original state without noticed inundation, conforms inundation expected from near offshore thrust fault.

First, we constructed a grid of the SRTM Lebanon topography merged with the EMODnet bathymetry of the northeastern Mediterranean Basin, and modified the present-day Beirut coastline so as to reflect its pattern at the time. We then modelled the coseismic deformation of an M7.5 thrust earthquake on the MLT, constraining the vertical offset according to evidence of uplifted marine-cut terraces along the Lebanese coast. The calculated seafloor deformation was used for tsunami wave generation, and non-linear shallow water equation for numerical modelling of tsunami propagation and inundation.

Preliminary assessment shows that, as expected, the simulated scenario exhibits a series of waves. However, the general effect of the simulation is a notable drawdown and minimal inundation, which in our eyes is compatible with the historical observations. The results also suggest that the modelled M7.5 MLT offshore scenario, can explain the 551 AD tsunami description with no need to consider secondary submarine and/or subaerial landslide sources. The review of historical events is thus an important tool to characterize earthquake and tsunami hazards in this area. While further elaboration is certainly needed, we already learnt the need to consider coseismic deformation in tsunami inundation modelling. This effect is critical in the case of near-shore sources leading to coseismic subsidence of coastal areas, which in turn can amplify the expected inundation.

How to cite: Salamon, A., Omira, R., and Baptista, M. A.: Modern Eyes on the Historical 551 AD Earthquake and Tsunami Offshore Phoenicia, Lebanon of Today, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4840, https://doi.org/10.5194/egusphere-egu22-4840, 2022.

09:33–09:40
|
EGU22-166
|
ECS
|
Virtual presentation
Elizaveta Tsukanova

The last major events in the Sea of Japan were in 1983 and 1993. There were the 1983 Nihonkai-Chubu Earthquake (Mw 7.8) and the 1993 Hokkaido Nansei-Oki Earthquake (Mw 7.7). These earthquakes caused tsunamis, which we are studying in this research. I use numerical modelling to reproduce and study effects for the Russian coast. The tsunami waves were stimulated by the TUNAMI numerical model. The bottom topography was created using GEBCO database (30 arc seconds), SRTM data, digitized Russian navigational charts and NOAA Center data. The tsunami source was calculated using Okada's formulas. To better resolve local resonant properties arising from local topography and tsunami run-up, calculations were carried out with nested grids. Using nested grids made it possible to obtain significant agreement with the observational data. Since the seismic source of the 1993 earthquake has a complex structure, three different models were analyzed: USGS, Harvard-model and Takahashi et al. 1995. This study focuses on an examination of the Russian coast. Vladivostok, Posyet and Nakhodka were considered in the most detail. Comparison of the model with the observations was done for both the tsunami waveforms and their spectra. Also, a tsunami wave height map was built for the entire Russian coast of the Sea of Japan. The maximum tsunami wave height on the Russian coast in 1993 was more than 5 m.

How to cite: Tsukanova, E.: The 1983 and 1993 tsunamis on the coast of the Sea of Japan: observations and numerical modelling, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-166, https://doi.org/10.5194/egusphere-egu22-166, 2022.

09:40–09:47
|
EGU22-904
|
On-site presentation
Jean Roger et al.

On the 12th of August 2021 at 18:32:54 and 18:35:20 (UTC) a doublet of reverse faulting earthquakes of magnitude Mw 7.5 and 8.1 were recorded by seismic observatories. These earthquakes were located on the South Sandwich Islands (UK) subduction zone, in the south Atlantic Ocean at 25.032°W/57.567°S and 25.327°W/58.451°S respectively (USGS locations). Initially, their temporal proximity (2’26”) made clear distinction of the two events impossible and a tsunami warning was issued by the PTWC after the first earthquake only. In fact, a tsunami was clearly recorded ~800 km north-westward of the epicentre on nearby King Edward Point coastal gauge (South Georgia Island, UK) ~1.5 hours after the shaking, showing a maximum amplitude of ~74 cm. While tsunami waves were recorded by neighbouring gauges located in the south Atlantic Ocean and the south-west Indian Ocean, numerical simulations of wave propagation show that this tsunami appears likely to have reached far-field regions not only in the Atlantic Ocean, but also in the Indian and Pacific Oceans using oceanic ridges like the Mid-Atlantic and Atlantic-Indian ridges as waveguides. Analysis of 33 records from gauges located within the maximum amplitude lobes of the simulated tsunami validates the modelling and the nearly worldwide spread of this tsunami. Further tsunami simulations using high-resolution nested grids to refine the bathymetry around the gauges (e.g. La Réunion Island, Cocos, Hillary Harbour) are used to constrain the source model via tsunami waveform inversion, comparing the calculated results and the real records. Consequently, we highlight that this tsunami reached many places including the Canary Islands, Cape Verde and the Azores in the northern Atlantic Ocean, and French Polynesia, New Zealand, Hawaii and as far as the Aleutian Islands in the Pacific Ocean, making this subduction zone a source for further consideration in tsunami hazard assessments of these distant regions, especially in the case of a more energetic rupture. Although the largest known event in the instrumental period is the 27 June 1929 MPAS 8.3 earthquake, geological knowledge of the region suggests that this ~1000 km long convergence zone between the South American and the South Sandwich plates with a convergence rate of 69-78 mm yr−1, is potentially able to produce a Mw 9.0 earthquake. This is supported by recent studies showing that the sediment thickness of 2-3 km at the trench and the ~150 km wide subduction interface shallow dipping (< 20° in the forearc part) are positive factors for generation of earthquakes Mw > 8.5. Results of simulation of Mw 9.0+ scenarios rupturing most of the subduction zone are discussed as well as the particular role of the oceanic ridges in the tsunami propagation. Our research aims to improve understanding of tsunami hazard posed by this subduction zone, especially for southern hemisphere coastlines.

How to cite: Roger, J., Hebert, H., Jamelot, A., Gusman, A., Power, W., and Hubbard, J.: The South Sandwich circum-Antarctic tsunami of August 12, 2021: widespread propagation using oceanic ridges, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-904, https://doi.org/10.5194/egusphere-egu22-904, 2022.

09:47–09:54
|
EGU22-6347
|
Virtual presentation
Jihwan Kim et al.

Two hazardous storms, Christina (January 2014) and Leslie (October 2018), destructively affected the coast of Portugal and generated extreme sea level variations. We analyzed both the sea-level and meteorological data, and performed numerical simulations to examine the observed wave-induced coastal hazard and identify the background harbor resonances at each port. The results revealed that the sea-level variation is affected by the combined effect of low-frequency sea level rise (surges) and high-frequency (HF) waves. For the 2014 event, we found that wind was the main source of the HF sea surface variation, which excited the background harbor resonance. For the 2018 event, storm surges were significantly stronger and HF amplitudes were mostly induced by the movement of a pressure jump, leading to a meteotsunami formation. Commonly, wind is considered as a principal factor of the storm-generated HF waves, but we show herein  that the atmospheric pressure jump can play an important role in their formation through meteotsunami. The latter, when combined to a storm surge, can cause serious impact on the threatened coastal areas. 

How to cite: Kim, J., Omira, R., and Dutsch, C.: Combined storm and meteotsunami hazards: Data analysis and numerical simulation of Christina (Jan. 2014) and Leslie (Oct. 2018) events on the coast of Portugal, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6347, https://doi.org/10.5194/egusphere-egu22-6347, 2022.

Wed, 25 May, 10:20–11:50

Chairpersons: Alexander Rabinovich, Jadranka Sepic

10:20–10:27
|
EGU22-10966
Alexander Rabinovich et al.

From 12 to 16 October 2016, a series of three strong low-pressure systems, including typhoon “Songda”, passed over the coasts of southern British Columbia (BC) and Washington State (WA). Typhoon “Songda” was generated on 2 October about 1,000 miles to the southwest of Hawaii. After passing along the coast of Japan, it turned eastward, crossed the Pacific Ocean, arriving off the coast of North America on 12 October, where it merged with local extratropical cyclones propagating along the coast of Vancouver Island.  These three lows passed across the western coast of the island on 14-15 October, generating strong surface currents if the offshore region and significant sea level oscillations, including storm surges, seiches and infragravity waves along southern BC and northern Washington. Oceanic observations of the event included HF WERA radar data, offshore bottom sea pressure measurements from the Ocean Network Canada (ONC) observatories and sea level records from BC and WA tide gauges. Meteorological data analyzed included radar records, satellite imaginary, reanalysis synoptic data, and air pressure and wind surface measurements of remarkable spatial and temporal resolution from more than 150 school network stations. These extensive datasets allowed for a detailed tracking of atmospheric processes responsible for strong ocean surface currents and sea-level oscillations. Maximum currents of up to 50 cm/s were measured by the HF radar. The surge heights on the southern BC and northern WA coasts were higher than 80 cm, with maximum storm surge observed at La Push, WA (117 cm) and New Westminster, BC (101 cm). A particularly interesting phenomenon was observed on the west side of Vancouver Island, beginning at Tofino, where the tide gauge record indicated a sharp, knife-like 40-cm increase in sea level with a peak value at 07:01 UTC on 14 October. Slightly lower sharp sea level peaks were also observed at Bamfield, Port Alberni and Port Renfrew. The high negative correlation between sea level and atmospheric pressure is consistent with the inverted barometer (IB) effect. Sharp sea level peaks at Tofino, Bamfield and Port Alberni are shown to be related to the specific shapes of the air pressure variations at these sites (the minimum atmospheric pressure at Tofino was 971.4 hPa), but the sea level response was 1.5-2.5 times greater than the IB effect, demonstrating the topographic amplification of sea levels in the respective areas. Such oscillations at Tofino and surrounding regions, may be described as a “meteorological tsunami” that for this specific case has a character of a forced solitary wave.

How to cite: Rabinovich, A., Šepić, J., and Thomson, R.: The extreme sea-level event of 14-15 October 2016 on the coasts of British Columbia and Washington State caused by Typhoon "Songda", EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10966, https://doi.org/10.5194/egusphere-egu22-10966, 2022.

10:27–10:34
|
EGU22-10261
|
Virtual presentation
Finn Løvholt et al.

On June 17, 2017, a 40 Mm3 rock avalanche generated a tsunami that impacted several coastal communities in Karrat Fjord, Central West Greenland. The tsunami run-up was 10-12 m high in the nearest village 30 kilometres away from the rock avalanche and caused four fatalities. The two villages most heavily affected are still evacuated. In the aftermath of this event, several unstable rock slopes have been discovered proximal to the 2017 rock avalanche. One of these volumes, coined Karrat 1, has a volume of about 0.5 km3 and is hence at least an order of magnitude larger than the volume involved in the 2017 event. To put this in perspective, it has a volume 2-3 times larger than the 2018 Anak Krakatau tsunami that led to more than 400 fatalities in Sunda Strait, Indonesia (which is also much more heavily populated). Hence, the Karrat 1 worst case scenario poses a threat to a much larger area than the event that took place in 2017 and could potentially affect the whole fjord system. In this study, we quantify the tsunami hazard from this unstable rock slope as well as the 2017 event. We first provide a set of landslide tsunami simulations using a frictional-collisional Voellmy type model coupled to a tsunamis model for the event in 2017 and compare it with observations. We found that the model results agree closely with observations of the tsunami run-up heights, observations of the tsunami arrival times, and the wave periods. The 2017 tsunami model was then used to calibrate the landslide source model for the future hazard, simulating the Karrat 1 landslide tsunami with an included uncertainty range. Extreme run-up heights (10-70 m) are found for the nearest villages, as well as complete inundation of entire low-lying villages, some more than 100 km away from the landslide release area. The large modelled run-up heights, involving extreme run-up heights and relatively short arrival times for the nearby villages, demonstrate the need for better understanding of the risk as well as risk-reducing measures. With few or no previous subaerial events that have taken place historically of this scale, the possible implications of a catastrophic release are widespread, but they also imply substantial uncertainties.

How to cite: Løvholt, F., Glimsdal, S., Harbitz, C., Svennevig, K., Keiding, M., and Møller, J. J.: Tsunami propagation and high-resolution inundation modelling of the 2017 Karrat rock avalanche and potential future tsunamis from proximal slope failures, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10261, https://doi.org/10.5194/egusphere-egu22-10261, 2022.

10:34–10:41
|
EGU22-11303
Tatok Yatimantoro et al.

A tsunami numerical inundation modeling in the Ambon city was developed by considering large earthquakes along the Ambon bay strike-slip fault and triggering submarine landslide as the tsunami source. 
The simulation was conducted using Comcot (Cornell Multi-grid Coupled Tsunami model) with a nested grid system in the spherical coordinate system. The four different spatial grid sizes of 60 (layer 1), 15 (layer 2), 3.75 (layer 3), and 0.9375 (layer 4) arc-sec were used in the computation. The linear shallow-water theory with bottom friction was applied for layers 1 -3, meanwhile, layer 4 used the non-linear shallow-water theory with manning roughness coefficient and detail bathymetry data. 
The single segmentation of earthquake scenarios with magnitudes Mw 7.2 was assumed. The earthquake then triggers submarine landslides in some areas around Ambon city. The landslide area was approached by Peak Ground Acceleration (PGA) value and historical data.
The results showed that in Ambon city the first tsunami wave arrived 18 min after the earthquake with a maximum flow depth of 7.4 m and inundation distance around 1.2 km. These results show that Ambon city has a risk of tsunami threat from earthquakes and submarine landslides. Therefore, it is necessary the tsunami hazard preparedness by the government and communities.

How to cite: Yatimantoro, T., Harvan, M., Anugrah, S. D., Daryono, D., Prayitno, B. S., and Adi, S. P.: Numerical Tsunami Inundation Modeling in Ambon City, Indonesia for Potential Earthquake and Landslide at Ambon bay, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11303, https://doi.org/10.5194/egusphere-egu22-11303, 2022.

10:41–10:48
|
EGU22-6439
|
Virtual presentation
Alain Rabaute et al.

Historical earthquake records suggest that the Alboran Sea seismicity is mostly triggered by strike-slip faults with little or no vertical throw preventing significant tsunami formation. Although in the North Alboran Sea the Averroes fault may have a tsunamigenic potential, the main active fault system responsible of the last three major earthquakes (Mw ≥ 6) in the South Alboran Sea, the Al-Idrissi fault, has no significant vertical component. This points to submarine landslides as the main potential source of tsunamis for the southern sector of the basin. Our study deals with the tsunamigenic potential of submarine landslides in the southern Alboran Sea, where several deposits are stacked within the last million year of sedimentary cover. We have identified up to 67 landslide events with volumes between 0.01 to 15 km3. The probability of landslide occurrence has been analysed with a logistic regression describing the relationship between a binary response variable (existence or absence of landslide) and a set of predictor variables such as high seafloor gradients and presence of active faults. The analysis of the severity of a given landslide has been investigated based on the estimation of the probability that the landslide reaches a certain (high) level (e.g. tsunami run-up or submarine cable breaks) giving that it has occurred through the extreme value analysis. We have used the Shaltop code simulating landslide run-out on the basis of a depth-averaged model based on the hydrostatic Saint Venant equations and Coulomb-type basal friction considering a Bingham rheology. Our tsunami simulations include Shaltop output scenarios as a source of the generated tsunami through hydrodynamic simulations using the hydrostatic 3D Navier-Stokes code Freshkiss3d. We found that tsunamis waves triggered by submarine landslides on the South Alboran Sea would be no higher than two meters. However, the tsunami would include wavelengths of tens of kilometres translating into important water volumes flooding several areas of around the Alboran coast. 

How to cite: Rabaute, A., Lafuerza, S., Thomas, M., Sainte-Marie, J., El Baz, A., Mangeney, A., d'Acremont, E., Basquin, E., Mercier, D., Creach, A., and Gorini, C.: Tsunami hazard along the Alboran Coast triggered by submarine landslides, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6439, https://doi.org/10.5194/egusphere-egu22-6439, 2022.

10:48–10:55
|
EGU22-1559
|
Virtual presentation
Mohammad Heidarzadeh et al.

Eastern Indonesia is exposed to significant tsunami hazards induced by its complex tectonic setting characterized by several curved subduction zones, multiple active volcanoes, as well as submarine landslides. Therefore, the region experiences tsunami from various types of sources (earthquake, landslide and volcano). Here, we study the great tsunami hazards in Eastern Indonesia through analyzing two recent real tsunamis that occurred in this region namely the 14 November 2019 Molucca Sea tsunami following an Mw 7.2 earthquake, and the 16th of June 2021 tsunami following an Mw 5.9 earthquake.

For the 2019 Molucca Sea tsunami, we analyzed 16 tide gauge records and 69 teleseismic data to characterize the tsunami and the earthquake. The maximum zero-to-crest tsunami amplitude was 13.6 cm recorded at Bitung. A combination of aftershocks analysis, forward tsunami simulations and teleseismic inversions were applied to obtain the tsunami source. It is found that the best results are obtained using a rupture velocity of 2.0 km/s and a high-angle reverse fault with a dip angle of 55o. The source model has a maximum slip of 2.9 m, and an average slip of 0.64 m. The seismic moment associated with this final slip model is 7.64 × 1019 N·m, equivalent to Mw 7.2. By comparing the results with other similar events in the region, such as the November 2014 event (Mw 7.1) with a reverse mechanism and a high dip angle of 65o, we may conclude that the Molucca Sea region is prone to splay faulting.

The 16th June 2021 tsunami was observed on the southern coast of Seram Island following an Mw 5.9 earthquake. The tsunami’s maximum wave amplitude was approximately 50 cm on the Tehoru tide gauge whereas the other two nearby stations showed amplitudes of less than 4 cm. Such a relatively large tsunami (50 cm in Tehoru) is normally unexpected from an earthquake of Mw 5.9 having a normal faulting mechanism. It is likely that a plausible secondary tsunami source, such as a submarine landslide, was involved. For the case of the 2021 Seram tsunami, here we apply numerical modelling and bathymetric analysis to examine the veracity of it being generated by a submarine landslide. Modeling of earthquake sources of the tsunami confirmed that that the simulated tsunamis were only a few centimeters in height and thus cannot reproduce the 50 cm waves observed in Tehoru. However, we were able to reproduce the tsunami observations using potential landslide sources.

This research is funded by The Royal Society (the United Kingdom), grant number CHL/R1/180173.   

How to cite: Heidarzadeh, M., Hilmann Natawidjaja, D., Hananto, N. D., Kongko, W., Sabeti, R., Daryono, M. R., Putra, P., Patria, A., and Gusman, A. R.: Tsunami hazards in Eastern Indonesia from earthquake, landslide and volcanic sources: Seram Island (June 2021) and Molucca Sea (November 2019) tsunamis, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1559, https://doi.org/10.5194/egusphere-egu22-1559, 2022.

10:55–11:02
|
EGU22-2852
|
Virtual presentation
Anita Grezio et al.

The coasts of the Mediterranean Sea are densely populated and exposed to tsunami inundations as reported by historical evidence. Measures to mitigate the tsunami risk in this region are based on Probabilistic Tsunami Hazard Assessments (PTHA) computed considering present coastal morphologies. However, mean sea level projections for the 21st century indicated a general sea level rise which can be substantially modified if uplift or subsidence may occur locally due to other geological factors. In order to reduce the potential impact of tsunamis all factors (climatic or not) should be included in the tsunami hazard analysis. In this study we focus on the Eastern Sicily and we examine how the PTHA can significantly change when the general trend of sea level rise, based on AR-5 and AR-6 IPCC climate scenarios and rates of Vertical Land Movements, are included in the region. Moreover, we take into account associated epistemic uncertainties related to the future sea level rise under different conditions of low- and high-emission representative concentrations. 

How to cite: Grezio, A., Baglione, E., Selva, J., Tonini, R., Anzidei, M., and Vecchio, A.: Probabilistic Tsunami Hazard Assessments in Eastern Sicily (Italy) including sea level rise caused by climate change and local subduction effects. , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2852, https://doi.org/10.5194/egusphere-egu22-2852, 2022.

11:02–11:09
|
EGU22-3949
|
ECS
|
Virtual presentation
Enrico Baglione et al.

We present a refined methodological procedure for computationally efficient local SPTHA based on regional SPTHA.  The adopted procedure extracts from the regional SPTHA the most impacting tsunami sources at the investigated site, and reconstructs hazard curves on high-resolution topobathymetric models based on a reduced set of inundation simulations. This procedure enhances the original workflow for local SPTHA quantification described by Volpe et al. (2019), applying some significant upgrades to simplify its application and improve the accuracy of the results. In particular, the description of local sources has been refined through a more detailed discretization of the natural variability (aleatory uncertainty), eventually reducing the epistemic uncertainty. Then, a more efficient filtering procedure, based on the strategy proposed by Williamson et al. (2020), is adopted to select a subset of scenarios to be modelled at high resolution, eventually reducing the epistemic uncertainty introduced by this selection. This allows to perform only coarse-grid simulations after the regional source filtering and local source refinement, and then combine coarse-grid results with fine-grid topography. Overall, the resulting method simplifies the original one, improving accuracy and decreasing uncertainty. The newly developed procedure is applied to an illustrative case study for the harbour of Ravenna (Northern Adriatic Sea, Italy).

How to cite: Baglione, E., Brizuela, B., Volpe, M., Armigliato, A., Zaniboni, F., Tonini, R., and Selva, J.:  An improved workflow to efficiently compute local seismic probabilistic tsunami analysis (SPTHA): a case study for the harbour of Ravenna (Italy), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3949, https://doi.org/10.5194/egusphere-egu22-3949, 2022.

11:09–11:16
|
EGU22-5642
|
ECS
|
|
On-site presentation
Naveen Ragu Ramalingam et al.

This work investigates a novel approach combining numerical modelling and machine learning, aimed at developing an efficient procedure that can be used for large scale tsunami hazard and risk studies. Probabilistic tsunami hazard and risk assessment are vital tools to understand the risk of tsunami and mitigate its impact, guiding the risk reduction and transfer activities. Such large-scale probabilistic tsunami hazard and risk assessment require many numerically intensive simulations of the possible tsunami events, involving the tsunami phases of generation, wave propagation and inundation on the coast, which are not always feasible without large computational resources like HPCs. In order to undertake such regional PTHA for a larger proportion of the coast, we need to develop concepts and algorithms for reducing the number of events simulated and more rapidly approximate the simulation results needed. This case study for a coastal region of Japan utilizes a limited number of tsunami simulations from submarine earthquakes along the subduction interface to generate a wave propagation database at different depths, and fits these simulation results to a machine learning model to predict the water depth or velocity of the tsunami wave at the coast. Such a hybrid ML-physical model can be further coupled with an inundation scheme to compute the probabilistic tsunami hazard and risk for the onshore region.

How to cite: Ragu Ramalingam, N., Johnson, K., Pagani, M., and Martina, M.: A hybrid ML-physical modelling approach for efficient approximation of tsunami waves at the coast for probabilistic tsunami hazard assessment, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5642, https://doi.org/10.5194/egusphere-egu22-5642, 2022.

11:16–11:23
|
EGU22-7194
|
Virtual presentation
Chiara Visentin et al.

Meteotsunamis (or meteorological tsunamis) are long, progressive sea waves triggered by external forcings due to meteorological events as e.g., air pressure disturbances, wind gusts and fast-moving storms that are observed in beaches of enclosed basins and/or in ocean waves entering the harbors and bays. The atmospheric disturbance in open sea generates near the surface water the localized waves, that travel at the same speed but with a period ranging from a few minutes to two hours. The waves propagate toward the shore amplifying near the coast due to resonance mechanisms related to the bathymetric characteristics of the waterbody and the topography of the coastal line. Therefore, a meteotsunamis results from two resonance effects: an external resonance between the air pressure disturbance and the long sea waves in the open sea, followed by an internal resonance between the incoming long waves and the harbor/bay eigenmodes.

Meteotsunamis have been observed all around the globe, but the most destructive events happened at a limited number of sites where meteorological and resonance conditions (i.e., intense resonant amplification due to the harbor/bay geomorphology, dynamic instability, frontal passages, gales, squalls, storms, tornadoes, convection cells, and atmospheric gravity waves) are satisfied at the same time. Examples of these sites are the North-East Adriatic Sea, the Balearic Islands (Spain) and the Sicily Strait (Marrobbio). Over the years, this natural phenomenon recorded an increase (higher frequency of Medicanes) and it has caused structural damages to properties and infrastructures along the coastal areas, as well as human casualties.

In the last fifteen years, numerous studies have addressed the issue of producing statistics and hazard estimates for meteotsunamis, even though in situ data are scarce and often available with a low spatial and temporal resolution. Numerical atmospheric-ocean models, mostly running with simulated air-pressure disturbance and calibrated over data of real events, were therefore carried out seeking to establish a shared approach for hazard estimation and meteotsunamis short-term forecast. Selecting appropriate models for this natural phenomenon is important in the view of planning coastal intervention in danger areas and quantifying the hazard in the harbor/bay in relation to geomorphological changes. In this light the PMO-GATE project (Preventing, Managing and Overcoming Natural-Hazards Risks to mitiGATE economic and social impact project) in the framework of the Interreg V Italy-Croatia 2014-2020 Program aims to develop a joint innovative methodology to strengthen and consolidate the collaboration against natural disasters specific to the NUTS Italy-Croatia in order to increase the level of protection, resilience and prevention of natural disasters through shared management methodologies and multi-risk overcoming of extreme events, such as meteotsunamis, to deal with natural risk with greater awareness and effectiveness.

In particular, it is crucial to understand whether and how the hazard estimate would be modified due to coastal changes brought about by the rise in the sea level expected as a consequence of climate changes.

How to cite: Visentin, C., Prodi, N., Benvenuti, E., Marrocchino, E., and Vaccaro, C.: Meteotsunamis: the hazard in the coastal areas, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7194, https://doi.org/10.5194/egusphere-egu22-7194, 2022.

11:23–11:30
|
EGU22-7972
|
On-site presentation
Lyubka Pashova* et al.

Tsunamis are severe natural hazards, causing significant human casualties and material damage to infrastructure, especially in the coastal zone. Research shows that tsunami danger exists for any water basin. The Black Sea is an inland sea, surrounded and crossed by several active faults whose geodynamic characteristics indicate that they can generate a tsunami. Moreover, the Black Sea is also prone to landslide-generated tsunamis and meteotsunamis. Until five decades ago, the existence of a tsunami threat in the Black Sea was ignored until the appearance of books that mention events described by ancient chroniclers interpreting information about tsunami-related phenomena in historical documents.

This work reviews and systematizes the main achievements in the field of tsunami research in Bulgaria from the initial voluntary enthusiastic research, initiated through the FP4-ENV 2C funded project "Genesis and impact of the tsunami on the European coasts" (GITEC-TWO, 1996-1998; https://cordis.europa.eu/project/id/ENV4960297) up to the present days. The small number of tsunami events observed in the western Black Sea basin limits our knowledge of the tsunamigenic potential of the Black Sea. The main problems, omissions and challenges are related to establishing the characteristics of tsunami sources, such as kinematic parameters of active faults and their geometry, coastal and underwater landslides and special weather conditions inducing meteotsunamis. This review presents the actions, studies, and observations on the western Black Sea coast, the first steps in building a tsunami warning system and other related activities. Based on the collected information, we identify the research gaps according to the AGITHAR priority matrix (Behrens et al., 2021) and highlight the emerging research areas in the Black Sea basin. The possibility of proposing a framework for assessing multi-hazard and multi-risk due to the cascade effect of different hazards along the Bulgarian coast in the context of the Sendai Framework for Disaster Risk Reduction is also outlined.

Acknowledgements: The authors thank the Bulgarian National Science Fund for co-funding the research under the Contract КП-СЕ-КОСТ/8, 25.09.2020, which is carried out within the framework of COST Action 18109 “Accelerating Global science In Tsunami HAzard and Risk analysis” (AGITHAR; https://www.agithar.uni-hamburg.de/).

 

References:

Behrens J, Løvholt F, Jalayer F, Lorito S, Salgado-Gálvez MA, Sørensen M, Abadie S, Aguirre-Ayerbe I, Aniel-Quiroga I, Babeyko A, Baiguera M, Basili R, Belliazzi S, Grezio A, Johnson K,Murphy S, Paris R, Rafliana I, De Risi R,Rossetto T, Selva J, Taroni M,Del Zoppo M, Armigliato A, Bures V, Cech P, Cecioni C, Christodoulides P, Davies G, Dias F, Bayraktar HB, González M, Gritsevich M, Guillas S, Harbitz CB, Kanoglu U, Macías J, Papadopoulos GA, Polet J, Romano F, Salamon A, Scala A, Stepinac M, Tappin DR, Thio HK, Tonini R, Triantafyllou I, Ulrich T, Varini E, Volpe M and Vyhmeister E (2021) Probabilistic Tsunami Hazard and Risk Analysis: A Review of Research Gaps. Front. Earth Sci. 9:628772. doi: 10.3389/feart.2021.628772

* corresponding author

How to cite: Pashova*, L., Didenkulova, I., and Ranguelov, B.: Tsunami research in Bulgaria: recent developments, gaps and further directions, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7972, https://doi.org/10.5194/egusphere-egu22-7972, 2022.

11:30–11:37
|
EGU22-9174
|
On-site presentation
Reneta Raykova and Lyuba Dimova

The Black Sea is located in the Anatolian sector of the Alpine-Himalayan orogenic system. In this region the African and Arab plates are moving to the north and to the west colliding with the Eurasia tectonic plate. In this study we focused on the northern Bulgarian Black Sea coast, where devastating earthquakes occurred in the past, during the Ist century BC, 543 AD, 1444 and 1901, all of them with estimated magnitudes M>7.0 causing tsunami waves. An evaluation of the possible seismic sources and maximum credible earthquake magnitude is made to build tsunami hazard scenarios for the northern Bulgarian coastline, including Shabla-Kaliakra seismic zone. The numerical code UBO-TSUFD is used for the tsunami simulations, coupled with bathymetry and relief data. The initial conditions of the generated tsunami waves are calculated using the method proposed by Okada supplemented with focal mechanisms information and fault geometry. We consider three seismic sources (SS I, SS II and SS III) which are tested for three different earthquake magnitudes M7.0, M7.5 and M8.0. To increase the resolution of the results we use nested grids, as the finest one (space resolution 50 m) is focused on the coastline between the city of Varna and Cape Kaliakra. We built simplified local tsunami hazard maps based on the computed water column on the coast for all nine tsunami scenarios in the studied region. The potentially threatened inundation zones are marked with different colors and vary between 0 and 5 m, depending on the selected magnitude. SS I poses the highest risk of potential tsunami flooding with the calculated water column for the northern part of the Bulgarian coast reaching more than 1.5 m, even for M7.0. When M7.5 is considered, the tsunami heights rise to 2.3 m and assuming M8.0, the water column exceed 4 m. The gulf of Bourgas is partially protected by Cape Emine, located to the north. It should be noted that the Romanian coast and more precisely the shores to the north of Constanta are seriously affected by the modelled scenarios, as the calculated inundation heights exceed 2.5 m for M8.0. The results for SS III show the lowest values of the vertical water column inland. The modeling estimates the sea level variations in certain points computing synthetic mareograms. Virtual mareograms near Varna, Balchik and Albena resort displays the evolution of the initiated tsunami heights in time. SS II and SS III have similar behavior for all three magnitudes. The dominant tsunamigenic source with extremely high waves is SS I.

In addition, the impact of these three seismic sources on the entire Black Sea coast is examined through the coarse grid of 500 m, the propagation field and the maximum computed tsunami heights.

This study is funded by the Bulgarian National Science Fund, grant number CP-06-COST-7/24.09.2020. LD contributed to the European Cooperation in Science and Technology COST project “AGITHAR-Accelerating Global science In Tsunami HAzard and Risk analysis”.

How to cite: Raykova, R. and Dimova, L.: Tsunami hazard scenarios for the northern Bulgarian Black Sea coast, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9174, https://doi.org/10.5194/egusphere-egu22-9174, 2022.

11:37–11:44
|
EGU22-6282
|
On-site presentation
Joern Behrens et al.

Probabilistic tsunami hazard and risk analysis (PTHA/PTRA) is an emerging scientific discipline within the tsuanmi community and allows potentially to incorporate the diverse sources of uncertainty into disaster prevention, preparedness, and mitigation activities. While there are a number of successful applications of this paradigm, it is still an emerging field with a number of unresolved research questions. 

In a collaborative effort members of the COST Action AGITHAR assessed the existing research gaps for PTHA/PTRA and identified almost 50 different topics worth of further research. An ad hoc expert judgement was conducted to weight these open questions with respect to their expected impact on the quality of the PTHA/PTRA results and their difficulty to be answered. The results of this collaborative effort will be reported highlighting the most challenging and most severe research gaps.

The presentation is based on the following publication:
J. Behrens, F. Løvholt, F. Jalayer, et al. (2021): Probabilistic Tsunami Hazard and Risk Analysis – A Review of Research Gaps, Frontiers in Earth Science, 9:114, DOI:10.3389/feart.2021.628772.

How to cite: Behrens, J., Løvholt, F., Jalayer, F., Lorito, S., Salgado-Gálvez, M. A., and Sørensen, M. and the AGITHAR Team: Assessing research gaps in probabilistic tsunami hazard and risk analysis, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6282, https://doi.org/10.5194/egusphere-egu22-6282, 2022.

Wed, 25 May, 13:20–14:50

Chairpersons: Hélène Hébert, Lyuba Dimova

13:20–13:27
|
EGU22-5324
|
ECS
|
|
On-site presentation
Edhah Munaibari et al.

As tsunamis propagate across open oceans, they remain largely unseen due to the lack of
adequate sensors. To help better mitigate the tsunami risk, we use a detection method that takes
advantage of the efficient coupling of tsunami waves with the atmosphere. Tsunami-induced
internal gravity waves thus travel upward in the atmosphere, where amplitude amplifies by several
orders of magnitude as the air density decreases with altitude. Once the waves reach the
ionosphere, they put charged particles into motion, creating propagative phenomena known as
Traveling Ionospheric Disturbances (TIDs). Thanks to the Global Navigation Satellites Systems
(GNSS), such disturbances can be monitored and observed using the Total Electron Content (TEC)
derived from the delay that the ionosphere imposes in the electromagnetic signals transmitted to
the Earth’s surface by the GNSS satellites. Here we show ionospheric TEC signatures following the
passage of three ocean-wide tsunami events: the two tsunamis triggered by the March 4th, 2021
8.1 Mw Kermadec Islands, New Zealand, and the July 29th, 2021 8.2 Mw Perryville, Alaska
earthquakes, as well as across the southern Atlantic following the tsunami generated by the
August 12th, 2021 8.1 Mw Sandwich Islands earthquake. We classify the observed TEC signatures
based on detection reliability and the potential connection to the tsunami wavefield. In addition,
we utilize an analytical model to investigate the source of these identified TEC signatures. Thus, we
ensure their gravity-waves origin and assess the characteristics (wavelength, period, etc.) of such
gravity waves, which is necessary to confirm they originate from the tsunami. Finally, to better
map the tsunami amplitude at the ocean level in various configurations, we examine, compare,
and contrast the amplitude of the identified tsunami-induced TEC signatures from geographically
sparse regions. We account for multiple parameters such as the local magnetic field, the azimuth,
and the distance to the tsunami source. They all affect the TEC signature detection and the
retrieval of the tsunami wavefield and, thus, potentially, the estimated risk.

How to cite: Munaibari, E., Rolland, L., Sladen, A., and Delouis, B.: Tsunami Ionospheric Monitoring Across the Pacific Ocean and the Southern Atlantic, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5324, https://doi.org/10.5194/egusphere-egu22-5324, 2022.

13:27–13:34
|
EGU22-10756
|
On-site presentation
Matthew Fouch et al.

Innovative deep ocean monitoring technologies are crucial to catalyzing fundamental improvements in mitigating natural disasters, reducing human vulnerabilities, and understanding environmental threats. An attractive but untapped resource is the global submarine fiber optic cable network, which carries over 95% of international internet traffic. Key components of undersea fiber optic cable systems are repeaters, which are placed every 60-100 km along the cable to provide optical signal amplification. Integrating environmental sensors, including seismic, pressure, and temperature sensors, would enable real-time data collection for environmental and infrastructure threat reduction, natural disaster mitigation, and cable system monitoring. 

A unique technology that will revolutionize the utility of these cables is the SMART (Sensor Monitoring And Reliable Telecommunications) cable concept. Although the concept has been evaluated for over 10 years by an international suite of agencies and institutions, developing a SMART repeater requires substantial investment in research and development to validate a technology that could transform an industry. To date, no commercial manufacturer has allocated the resources to produce a prototype SMART repeater. To bridge this gap, we have obtained support by the National Science Foundation’s Small Business Innovation Research (SBIR) program to develop a benchtop prototype SMART repeater. As part of an international effort to help develop a SMART Cable system for the New Caledonia - Vanuatu region, we also have received support from the Gordon and Betty Moore Foundation as part of a team led by the University of Hawai`i.

Best-in-class SMART repeater sensors include a 3-axis accelerometer, absolute pressure gauge, and temperature sensor. Included with the sensors are data acquisition circuits with suitable dynamic range and precision, integration around a common communications module, an interface suitable for fiber optic cable spans up to 120 km in length, the software and firmware necessary to support the data path from the sensors to data storage servers, and precision timing for both time-stamps and frequency reference. The SMART repeater sensor system design is modular, thereby containing branch points for different sensors, as well as incorporation in different repeater housings or as standalone units. 

SMART Cables will be particularly well suited for providing essential tsunami monitoring data, particularly from the seismic and pressure sensors. More specifically, SMART repeaters provide a unique opportunity to develop significantly more extensive sensor networks of real-time ocean bottom monitoring, filling in critical near-field and azimuthal gaps frequently encountered in earthquake monitoring. Further, our SMART repeater sensor system design includes the option for either acceleration or velocity monitoring, thereby enabling better measurement of amplitudes of tsunamigenic subduction zone earthquakes while providing a lower noise sensor in ocean basins. Further, data from SMART Cables will facilitate the detection of other tsunamigenic sources, including underwater landslides. We will present the results of our sensor development efforts and upcoming opportunities for SMART Cable installations.

How to cite: Fouch, M., Lentz, S., Howe, B., and Avenson, B.: SMART Cables: Integration of Environmental Sensors Into Submarine Telecommunications Cables for Improved Ocean Monitoring, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10756, https://doi.org/10.5194/egusphere-egu22-10756, 2022.

13:34–13:41
|
EGU22-8075
|
On-site presentation
Rui Barbara et al.

Seismic instrumentation is critical for instantaneous tsunami early warning systems as well as assessing long-term risk of tsunami activity in areas with high seismic hazard. Ocean Bottom Seismometer (OBS) systems provide real-time data in areas with appropriate infrastructure or batch data from offline temporary autonomous stations.

OBS systems detect ground motion from seismic waves significantly before detecting any pressure change in the water column from an associated tsunami due to the order of magnitude difference in wave velocity. Güralp’s OBS systems combine seismic and pressure detection in both permanent cabled networks and temporary non-cabled systems utilising near-real-time acoustic transmission. All seismic sensors used in Güralp systems are sensitive to both earthquakes as well as other tsunami-triggering events such as landslides (e.g. Anak Krakatau, 2018) or volcanic eruptions (e.g. Hunga Tonga–Hunga Haʻapai, 2022).

Cabled systems provide obvious benefits of real-time data, confidence of installation and flexibility to add additional instrumentation without power consideration. For example, Güralp Orcus and Maris cabled OBS systems are both deployed off the western coast of North America monitoring volcanic and tectonically induced earthquakes that have potential to cause tsunamis. Seismometers at these stations coupled with pressure gauges allow for immediate notification of a threat and subsequent refinement of hazard estimates using surrounding assets such as dedicated DART buoys.

Both Orcus and Maris allow for multiple auxiliary systems to be incorporated into the system while maintaining as well as providing additional installation flexibility for operators. Orcus has facility for both strong & weak motion seismometers in addition to auxiliary sensors while Maris has the unique feature of operating at any angle without the need for a gimbal mechanism, simplifying installation and network design considerations.

The Güralp Aquarius is the latest generation autonomous OBS for short-to-medium term or rapid response campaigns to monitor areas with increased seismic and tsunami hazard. Aquarius also uses omnidirectional capabilities as well as acoustic communication of seismic data to the surface to improve operator confidence of installation. Acoustic communication also allows for near-real-time communication with land-based warning systems after a significant seismic event in anticipation of a tsunami. This can be verified and communicated after the initial seismic wave using onboard pressure gauges. In areas where surface communication is not required, intelligent battery systems optimise deployment lengths beyond 18 months for maximum data/cost benefit.

Güralp is also pioneering the use of seismic sensors and auxiliary equipment within Science Monitoring And Reliable Telecommunications (“SMART”) cables which have already been shown to be useful in incorporating pressure gauges to detect tsunami events. These cables utilise regular telecommunication cables making uses of their natural communication and power source qualities to improve sensor network coverage. Güralp is currently manufacturing a demonstration system to be deployed in the Ionian Sea, monitoring seismic and volcanic activity with the aim of indicating practicality and data quality using this installation method.

How to cite: Barbara, R., Cilia, M., Reis, W., Watkiss, N., Mohr, S., Hill, P., and Whealing, D.: Utilising ocean bottom seismometer platforms for tsunami early warning and hazard assessment, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8075, https://doi.org/10.5194/egusphere-egu22-8075, 2022.

13:41–13:48
|
EGU22-10179
|
Virtual presentation
Patricio A. Catalan et al.

Tsunamis have the potential to cause widespread damage and loss of life over large swaths of coastal areas. To mitigate their effects, both in the long term and during emergency situations, an accurate, detailed and timely assessment of the hazard is essential. Here, an enhanced method for estimating tsunami time series using a uni-dimensional convolutional neural network model is presented, with the aim of reducing the time and computing capacity required by a high-resolution numerical modeling. While the use of deep learning for this problem is not new,  most of existing research has focused on the determination of the capability of a network to reproduce inundation values. However, for the context of Tsunami Early Warning, it is equally relevant to assess whether the networks can predict the absence of inundation. Hence, the network model was adjusted for the bays of Valparaíso, Viña del Mar and Coquimbo in central Chile, based on a set of 6800 scenarios with Mw 8.0-9.2. Tentative models were trained with time series from low- and high-resolution numerical modeling, to recreate the tsunami time series of control points on land. The objective was to reproduce the inundation high resolution time series, when the network was fed with low resolution offshore data. The approach considered 1075 (15%) scenarios to test the model, and 5783 (85%) scenarios to adjust (train and validate) the model. Different performance metrics are employed, particularly the RMSE measured with respect to peak flow depth and arrival times. Critically, the number of false alerts and alerts not issued was analyzed, which was considered a relevant performance owing to the wide range of magnitudes tested that led to an unbalance between scenarios that inundate and the ones that not. A notable outcome in this study shows the network is capable of reproducing inundation, either for small or large amplitudes, and also of no inundation. To further assess the performance, the model was tested with three existing tsunamis and compared with actual inundation metrics at three cities with different hydrodynamic response. The results obtained are promising, and the proposed model could become a reliable alternative for the calculation of tsunami intensity measures (TIMs) in a near to real time manner, with a network model forecasting where sea surface and geodetic data are not readily available, as occurs in many countries. This could complement existing early warning systems to reduce uncertainties involved in the decision making process.

How to cite: Catalan, P. A., Núñez, J., Valle, C., Zamora, N., and Valderrama, A.: Estimating Time Series of Tsunami Inundation using One-Dimensional Convolutional Neural Networks for Early Warning., EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10179, https://doi.org/10.5194/egusphere-egu22-10179, 2022.

13:48–13:55
|
EGU22-12716
|
ECS
|
On-site presentation
Pierre Andraud et al.

Tsunami warning systems currently focus on the first parameters of the earthquake, based on a 24-hour monitoring of earthquakes, seismic data processing (Magnitude, location), and tsunami risk modelling at basin scale.

The French Tsunami Warning Center (CENALT) runs actually two tsunami modelling tools where the water height at the coast is not calculated (i.e., Cassiopee based on a pre-computed database, and Calypso based on real time simulations at basin scale). A complete calculation up to the coastal impact all along the French Mediterranean or Atlantic coastline is incompatible with real time near field or regional forecast, as nonlinear models require fine topo-bathymetric data nearshore and indeed a considerable computation time (> 45 min). Predicting coastal flooding in real time is then a major challenge in near field context, the aim being a rapid determination of shoreline amplitude and real time estimation of run-up and currents. A rapid prediction of water heights at the coast by amplification laws or derived transfer function can be used to linearly approximate the amplitude at the coastline, with error bars on calculated values within a factor 2 at best. However, such approach suffers from a limited consideration of local effects and no run-up estimation.

The goal is there to add complexity to the predicted models through deep learning techniques, which are newly explored approaches for rapid tsunami forecasting. Several architectures, treatments and settings are being explored to quickly transform a deep ocean simulation result into a coastal flooding model. The models provide predictions of maximum height and run-up, maximum retreat, and currents in 1 second. However, such approach is dependent of a large scenario base for learning. This work presents preliminary comparisons of the coastal impact captured from nonlinear time consuming tsunami simulations (ground truth) with predicted localised tsunami responses provided by rapid forecasting deep learning approaches at 10 m resolution along the French Mediterranean, for several earthquake scenarios.

How to cite: Andraud, P., Gailler, A., Sprunck, T., and Vayatis, N.: Deep learning models  exploration for rapid forecasting of coastal tsunami impact in near field context – application to the French Mediterranean coastline., EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12716, https://doi.org/10.5194/egusphere-egu22-12716, 2022.

13:55–14:02
|
EGU22-11876
|
Highlight
|
Virtual presentation
Cecilia Valbonesi

One of the most critical part of tsunami warning systems is the so-called “last mile”, i.e., informing promptly residents and tourists about a possible impending inundation.

In Italy, one of the most recent activities to reach this goal is the implementation of the Tsunami Ready (TR) Program, developed under the aegis of UNESCO and achieved in synergy between INGV, ISPRA and the Italian Civil Protection Department (the three components of the Italian Tsunami Warning System - SiAM).

In 2020, the path towards the TR recognition has started in three Italian pilot municipalities: Minturno, Palmi, Marzamemi. The response of local authorities has been enthusiastic in all three cases, despite numerous bureaucratic obstacles to involvement and membership.

Italy as a NEAM member aims to reach the goal of 100% of communities at risk of tsunami prepared for and resilient to tsunamis by 2030 through the implementation of the UNESCO/IOC Tsunami Ready Programme.

Several developments are going on because all participants are aware that TR is a virtuous model for dealing with tsunami risk, with numerous implications in terms of education and responsibilities for the harmful consequences of a tsunami.

First of all, the direct involvement of citizens in the education and information process represents a significant step change of TR. It is achieved through the participation of citizens’ representatives in the TR Local Board, which is responsible for monitoring the development of procedures and certifying that a suite of 12 target parameters identified in the TR guidelines have been accomplished.

It is important to remind that the recognition as Tsunami Ready community must be also approved by the National TR Board and by the UNESCO ICG.

Secondly, the existence of internationally accredited guidelines (IOC UNESCO n. 74 and its ongoing updates) represents a reliable parameter for determining the behavior to be adopted by public decision-makers.  In case of harmful events, the compliance with these parameters can contribute to mitigating the (possible) criminal reproach against civil protection officers charged in risk management.

How to cite: Valbonesi, C.: Tsunami Ready Programme in NEAM region: strategies, responsibilities and further advancements to protect communities from tsunamis, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11876, https://doi.org/10.5194/egusphere-egu22-11876, 2022.

14:02–14:09
|
EGU22-8069
|
ECS
|
Highlight
|
|
On-site presentation
Rachel Hunt et al.

Individuals and communities are known to respond in different ways to official tsunami warnings and natural tsunami warning signs. This interdisciplinary research seeks to understand how official warnings are decided upon and communicated, and the ways in which warnings can be tailored through educational measures to improve tsunami awareness and preparedness. By improving the understanding of tsunami responses to official warnings and natural warning signs through examining the interactions between different emergency agencies, the mitigation methods for various tsunami hazards, and the numerous approaches to public warning communication, it is proposed that more tsunami resilient communities can be developed in New Zealand.

Online social research methods were used to investigate tsunami early warnings and responses in New Zealand. 106 documents and archives were collected to examine the nature and content of official tsunami information and the methods currently used to communicate these warnings, including director’s guidelines, memorandums of understanding, standard operating procedures, ministerial reviews, and technical standards. 57 semi-structured interviews were conducted with tsunami researchers, warning specialists, and emergency managers to gain an understanding of the opinions held on the effectiveness of official warnings and public education. The participants were recruited from research institutes, national agencies, regional groups, and local councils in New Zealand, Australia, the Pacific Islands, the UK, and the USA.

Three key findings have been established. First, the division of responsibilities between the various research institutes, national agencies, regional groups, and local councils involved in monitoring, disseminating, and responding to official tsunami warnings leads to the potential for error and delay in issuing official warnings, highlighting the need for consistent messages and coordinated responses. Second, whilst New Zealand has the capability to communicate official warnings for distal events, the country relies on educating the public to observe natural warning signs for local events, with emergency drills as well as awareness and preparedness campaigns in place to promote self-evacuation. Third, whilst sirens can be useful for issuing official tsunami warnings in rural or isolated communities, they can create confusion if the tone is misunderstood, whilst Emergency Mobile Alerts (EMAs) can only be used in areas with good reception but provide more information on the approaching hazard.

Further public education around the warning communications issued by national, regional, and local agencies, as well as New Zealand’s vulnerability to distally, regionally, and locally generated tsunamis, would contribute to more effective tsunami responses. The advantages and disadvantages of sirens and EMAs emphasise the value of these two methods of tsunami warning being used holistically, in a multi-channel approach, to provide more thorough warning communication. This research concludes that improvements must be made to emergency agency interaction, tsunami mitigation methods, and warning communication approaches in order to develop tsunami resilience in New Zealand.

How to cite: Hunt, R., Fearnley, C., Day, S., and Maslin, M.: The Role of Communication and Public Education in Tsunami Early Warnings and Responses in New Zealand, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8069, https://doi.org/10.5194/egusphere-egu22-8069, 2022.

14:09–14:16
|
EGU22-8374
|
Virtual presentation
Fatemeh Jalayer et al.

Methodology:

A fragility model expresses the probability of exceeding certain damage levels for a given level of intensity for a specific class of buildings or infrastructure. An empirical tsunami fragility curve for a given damage level is derived based on observed pairs of data for the tsunami intensity measure and the corresponding damage level. Tsunami inundation depth and/or flow velocity are usually adopted as scalar intensity measures (they can also be employed together as a vector-valued intensity measure). Physical damage levels are usually defined in a hierarchical manner, implying discrete, mutually exclusive, and collectively exhaustive (MECE) damage states. This means that the fragility curves for consecutive hierarchical damage levels must not intersect. It is clear that by fitting empirical fragility curves to each single damage level, this condition is not automatically satisfied. To overcome this problem, ordered (“parallel”) fragility models or partially ordered models have been adopted in the literature to derive fragility curves for MECE damage states. Empirical tsunami fragility curves are usually constructed using generalized linear regression models by adopting probit, logit, or the complementary loglog link functions. As far as model comparison and selection are concerned, established statistical approaches have been used in recent literature to identify the optimal link function among those mentioned above. Moreover, for estimating the uncertainty in the resulting empirical fragility curves, bootstrap resampling has been commonly used.

The present work proposes a simulation-based Bayesian method for inference and model class selection to perform ensemble modelling of the tsunami fragility curves for MECE damage states and the related uncertainties for a given class of buildings. The method uses adaptive Markov Chain Monte Carlo Simulation (MCMC), based on likelihood estimation using point-wise intensity values, to estimate the fragility model parameters and the uncertainties. Among the set of viable fragility models considered, Bayesian model class selection is used to identify the simplest model that fits the data best (i.e., is a parsimonious model). The proposed method provides consistent parameter estimation and confidence intervals for MECE the damage states and identifies the best fragility model class among the pool of viable models, based on a single set of simulation realizations. The whole procedure is provided as open-source software on the site of the European Tsunami Risk Service (https://eurotsunamirisk.org/software/) and is also available as a standalone docker application.

Application:

As the case-study application, the central South Pacific region-wide tsunami on September 29, 2009 is used. The tsunami was triggered by an unprecedented earthquake doublet (Mw 8.1 and Mw 8.0). The tsunami seriously impacted numerous locations in the central South Pacific. Herein, the damage data related to 120 brick masonry residential buildings associated with the reconnaissance survey sites of American Samoa and Samoa islands were utilized as a proof of concept. A six-tier damage scale is considered, and tsunami inundation depth has been used as the intensity measure.

 

Keywords: probabilistic tsunami risk analysis, tsunami fragility, Bayesian inference, model class selection

How to cite: Jalayer, F., Ebarahimian, H., Trevlopoulos, K., and Bradley, B.: Empirical tsunami fragility modelling for hierarchical damage levels: application to damage data of the 2009 South Pacific tsunami, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8374, https://doi.org/10.5194/egusphere-egu22-8374, 2022.

14:16–14:23
|
EGU22-1273
|
ECS
|
On-site presentation
Anjar Tri Laksono et al.

The tsunami that occurred on the Southern Coast of West Java and Central Java resulted in 802 people killed, 498 people injured, and 1623 houses heavily damaged. The total economic loss and damage to infrastructure due to this disaster reached US$55 million. The impact of this disaster in Jetis Village, Cilacap, Central Java was 12 people died, Jetis Beach tourist facilities were damaged, transportation infrastructure was destroyed, and hundreds of houses collapsed. The Jetis area and its surroundings are very close to vital national infrastructures such as the Cilacap steam power plant that supplies electricity to southern Java and the Cilacap container port. In addition, this area is a tourist attraction visited by thousands of people per year. Therefore, the purpose of this research is to create a tsunami disaster mitigation map and evacuation route in Jetis Village to anticipate future casualties and economic losses. The method used in this study is scoring to create a tsunami mitigation map and Dijkstra's algorithm to determine the fastest evacuation route. The results depict that there are five zones of tsunami vulnerability, namely high impact potential, moderately high, moderate, moderately low, and low impact potential. The most vulnerable tsunami is the South Jetis area that has low elevation, is near the coast, fairly gentle slope, and is close to the river. Meanwhile, the northern part of Jetis is the safest zone of tsunami hazard. It has a high elevation, far from the coastline and river, and a steep slope. The distance of the evacuation route from the high-impact zone to the safe evacuation zone is 683 m. This study concludes that the high-impact to moderate-impact zone needs to be avoided in the event of a tsunami. If the community is within that range zone, then an evacuation route should be followed.

How to cite: Laksono, A. T., Widagdo, A., Aditama, M. R., Fauzan, M. R., and Kovacs, J.: Tsunami Mitigation Map and Evacuation Route Modeling on the Jetis Beach, Cilacap Regency, Indonesia using Scoring Method and Dijkstra’s Algorithm, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1273, https://doi.org/10.5194/egusphere-egu22-1273, 2022.