Description: This study constitutes a preliminary assessment of probabilistic tsunami inundation in the NE Atlantic region. We developed an event-tree approach to calculate the likelihood of tsunami flood occurrence and exceedance of a specific near-shore wave height for a given exposure time. Only tsunamis of tectonic origin are considered here, taking into account local, regional, and far-field sources. The approach used here consists of an event-tree method that gathers probability models for seismic sources, tsunami numerical modeling, and statistical methods. It also includes a treatment of aleatoric uncertainties related to source location and tidal stage. Epistemic uncertainties are not addressed in this study. The methodology is applied to the coastal test-site of Sines located in the NE Atlantic coast of Portugal. We derive probabilistic high-resolution maximum wave amplitudes and flood distributions for the study test-site considering 100- and 500-year exposure times. We find that the probability that maximum wave amplitude exceeds 1 m somewhere along the Sines coasts reaches about 60 % for an exposure time of 100 years and is up to 97 % for an exposure time of 500 years. The probability of inundation occurrence (flow depth 〉0 m) varies between 10 % and 57 %, and from 20 % up to 95 % for 100- and 500-year exposure times, respectively. No validation has been performed here with historical tsunamis. This paper illustrates a methodology through a case study, which is not an operational assessment.

Description: Probabilistic Tsunami Hazard Assessment requires the existence of a synthetic earthquake catalogue obeying the geodynamic framework of the source area and a fault model for each event. Usually, we assume a constant fault displacement, eventually filtered by a simple closure condition. The distribution of events by magnitude classes follows a truncated Gutenberg-Richter law with epicentres uniformly distributed in the source domain. This approach makes that the slip on the source structures will extend beyond the generation domain being heterogeneous inside that domain. This contradicts the assumption that the slip is caused by geodynamic constraints. Here, we propose a methodology for creating synthetic earthquake catalogues and we apply it to the Gloria Fault in the Atlantic. With three instrumental tsunamis recorded: 08.05.1939, 25.11.1941 and 26.05.1975 [1]. Additionally, we propose a semi-empirical scaling law for the generation of the tsunamigenic earthquakes

Description: Highlights • Novel conceptional model for the tsunamigenesis of coastal mass-wasting events. • Unlocking the debate on the tsunamigenic potential of small-scale coastal collapses. • Benchmarking a cliff-failure tsunami numerical model with historical data. • First detailed investigation of the 1930 Cabo Girão tsunami in Madeira, NE Atlantic. Mass-wasting events are a key process in the evolution of volcanic ocean islands. They occur at various dimensional scales and present a major source of hazard. When the collapsed material plunges into the sea, destructive tsunamis can be generated. Yet, the hazard potential of collapse-induced tsunamis is still poorly understood with different opinions on what consequences to expect from this type of events, particularly those related to massive volcanic island flank collapses. In this paper, however, we explore the hazard extent of tsunamis triggered by the smaller – but more frequent – coastal cliff-failures, in order to isolate critical factors in the generation, propagation and impact of these tsunamis. To achieve this, we use the prime example of Madeira, a volcanic island in the Atlantic Ocean highly vulnerable to cliff-failure. Particularly, we explore the March 4th, 1930 Cabo Girão event that triggered a deadly tsunami. The coastal impact of the 1930 “Deadly Wave”, as the island's inhabitants referred to the generated tsunami, resulted in 19 fatalities. We use historical description, morphological analysis, and numerical modelling to better understand the tsunamigenesis of tall island cliffs failing into the sea. Interestingly, we find that a relatively small-scale mass-wasting event (∼0.003 km3 volume) was the cause of the reported tsunami that inundated the nearest coasts. Our numerical results, fairly agreeing with the available collapse and subsequent tsunami descriptions, suggest that the tsunami impact was mainly localized on the southern coast of Madeira Island. Furthermore, our study allows proposing a novel morphology-based conceptional model for the tsunamigenesis and hazard extent induced by mass-wasting events on oceanic volcanic islands.

Description: The method presented here aims to assess the tsunami threat very rapidly after the occurrence of a large earthquake, using as input the parameters of the seismic source, and an approach based on Green's summation. We show that the main weakness of the approach (the need to consider only linear shallow water propagation) is largely compensated by the advantages in terms of computing performance and independence with respect to pre-computed scenarios. To test the approach and to illustrate its implementation in a real environment, we focus on the Sea of Oman, a tsunamigenic area characterized by Makran subduction zone which detailed structure is partially unknown and where secondary tsunami sources must also be taken into account, both for hazard studies and warning purposes. The potential source area is partitioned into a grid of unity water sources. A shallow water (SW) numerical model is used to pre-compute the corresponding empirical Green's functions on several points of interest located on the coasts of Iran, Pakistan and Oman. The comparison between Green's summation and the direct SW computation using the full resolution of the bathymetric grid shows that the accuracy is good enough for practical applications.

Description: Remarkable accumulations of marine boulders located above present spring tide level occur in two coastal lowlands of the Algarve (Portugal). The size-interval of particles studied here is seldom reported in the literature in association with extreme events of coastal inundation, thus making this study of relevance to many other coasts worldwide. The spreads of boulders extend several hundred meters inland and well beyond the present landward limit of storm activity. The marine origin of the boulders is demonstrated by well-developed macro-bioerosion sculpturing and in situ skeletal remains of endolithic shallow marine bivalves. The good state preservation of the fossils within the boulders indicates that abrasion during transport and redeposition was not significant. We envisage boulder deposition as having taken place during the Lisbon tsunami of AD 1755 through the simultaneous landward entrainment of coarse particles from nearshore followed by rapid shoreward suspended-dominated transport and non-graded redeposition that excluded significant sorting by weight or boulder dimensions. We use numerical hydrodynamic modeling of tsunami (and storm) waves to test the observational data on boulder dimensions (density, size, distribution) on the most likely processes of sediment deposition. This work demonstrates the effectiveness of the study of boulder deposits in tsunami reconstruction. Copyright © 2011 John Wiley & Sons, Ltd.

Description: The NEAM Tsunami Hazard Model 2018 (NEAMTHM18) is a probabilistic hazard model for tsunamis generated by earthquakes. It covers the coastlines of the North-East Atlantic, the Mediterranean, and connected Seas (NEAM). In this online data product, the hazard results are provided by hazard curves calculated at 2,343 Points of Interest (POI), distributed in the North-East Atlantic (1,076 POIs), the Mediterranean Sea (1,130 POIs), and the Black Sea (137 POIs) at an average spacing of ~20 km. For each POI, hazard curves are given for the mean, 2nd, 16th, 50th, 84th, and 98th percentiles. Maps derived from hazard curves are Probability maps for Maximum Inundation Heights (MIH) of 1, 2, 5, 10, 20 meters; Hazard maps for Average Return Periods (ARP) of 500, 1,000, 2,500, 5,000, 10,000 years. For each map, precalculated displays are provided for the mean, the 16th percentile, and the 84th percentile. All data are also made accessible through an interactive web mapper and through Open Geospatial Consortium standard protocols. The model was prepared in the framework of the European Project TSUMAPS-NEAM (http://www.tsumaps-neam.eu/) funded by the mechanism of the European Civil Protection and Humanitarian Aid Operations (grant no. ECHO/SUB/2015/718568/PREV26).