Alberto Armigliato

Tsunamis are unpredictable and infrequent but potentially large impact natural disasters. To prepare, mitigate and prevent losses from tsunamis, probabilistic hazard and risk analysis methods have been developed and have proved useful. However, large gaps and uncertainties still exist and many steps in the assessment methods lack information, theoretical foundation, or commonly accepted methods. Moreover, applied methods have very different levels of maturity, from already advanced probabilistic tsunami hazard analysis for earthquake sources, to less mature probabilistic risk analysis. In this review we give an overview of the current state of probabilistic tsunami hazard and risk analysis. Identifying research gaps, we offer suggestions for future research directions. An extensive literature list allows for branching into diverse aspects of this scientific approach.

... According to documents provided by the Intergovernmental Coordination Group for the Indian Ocean Tsunami Warning and Mitigation System (ICG/IOTWS, 2007), a tsunami hazard scenario is built up by specifying the various characteristics of a tsunamigenic source. ...

Scenarios represent a very useful technique for the definition and evaluation of tsunami hazard and risk for any given region, and a basic step in the frame of tsunami mitigation and preparedness and of sustainable coastal zone development. With the exception of very few countries, like Japan and the United States, emergency plans in the rest of the world have never taken serious care of tsunamis until the occurrence of the giant Indian Ocean tsunami on December 26, 2004. That event dramatically brought the problem of tsunami hazard and risk assessment to the general attention and showed the urgent need for implementation of tsunami early warning systems (TEWSs). The problem is particularly urgent for the Mediterranean countries that are known to have been attacked by numerous tsunamis in the past, several of which had catastrophic size and impact. This paper is an attempt to develop some simple scenarios of earthquake-generated tsunamis in the Mediterranean. We identify four different seismogenic areas in the western, central and eastern sectors of the basin. For each of them, we take into account a seismic fault capable of generating an earthquake with magnitude equal or larger than the highest magnitude registered in that region in historical times. Then we simulate numerically the ensuing tsunamis, highlighting the basic features of the wave propagation and roughly identifying the coastal sectors that are expected to suffer the heaviest tsunami effects. One of the most important outcomes is that these scenario tsunamis attack the nearest coasts within at most 15 minutes, which poses serious constraints for designing appropriate TEWS for the Mediterranean.

Italy has been affected by large tsunamis in the past. From historical catalogues the occurrence rate of tsunamis in the Italian seas is about 15 events per century, which shows that tsunamis are very rare phenomena and that probabilistic techniques cannot be applied with confidence, especially if interest is not nation-wide but is focussed on regional coastal areas. Here a method is presented that derives tsunami potential from the assessment of the occurrence rate of tsunamigenic earthquakes, and that, therefore, makes use of seismic catalogues as the primary source of information. The method is restricted to tsunamis of seismic origin, and nothing can tell on tsunamis generated by volcanic activity and by mass movements. Improving a methodology originally used for a preliminary evaluation of tsunami hazard in Italy more than one decade ago (Tinti, 1991), this paper applies probabilistic seismic hazard techniques focussing on south-west Italy, namely on Calabria and Sicily, that are among the most active seismic regions in Italy. The analysis is based on the Italian seismic catalogue known as CPTI2, that was recently released (2004) and that is integrated with the INGV catalogue, spanning a time period longer than 2000 years. The main steps of the procedure are: 1) estimating the occurrence rate of tsunamigenic earthquakes; 2) assessing the initial disturbance of the sea, with the aid of appropriate relationships between the earthquake size and the ensuing tsunami size; 3) evaluating the expected maximum tsunami height on the coast, on the basis of the known propagation properties of tsunamis. As the result of the analysis, estimated return periods of earthquake-induced tsunamis capable of producing coastal wave heights exceeding a given threshold are computed and shown in the form of suitable maps.

ABSTRACT Geoscientists deal often with hazardous processes like earthquakes, volcanic eruptions, tsunamis, hurricanes, etc., and their research is aimed not only to a better understanding of the physical processes, but also to provide assessment of the space and temporal evolution of a given individual event (i.e. to provide short-term prediction) and of the expected evolution of a group of events (i.e. to provide statistical estimates referred to a given return period, and a given geographical area). One of the main issues of any scientific method is how to cope with measurement errors, a topic which in case of forecast of ongoing or of future events translates into how to deal with forecast uncertainties. In general, the more data are available and processed to make a prediction, the more accurate the prediction is expected to be if the scientific approach is sound, and the smaller the associated uncertainties are. However, there are several important cases where assessment is to be made with insufficient data or insufficient time for processing, which leads to large uncertainties. Two examples can be given taken from tsunami science, since tsunamis are rare events that may have destructive power and very large impact. One example is the case of warning for a tsunami generated by a near-coast earthquake, which is an issue at the focus of the European funded project NearToWarn. Warning has to be launched before tsunami hits the coast, that is in a few minutes after its generation. This may imply that data collected in such a short time are not yet enough for an accurate evaluation, also because the implemented monitoring system (if any) could be inadequate (f.i. one reason of inadequacy could be that implementing a dense instrumental network could be judged too expensive for rare events) The second case is the long term prevention from tsunami strikes. Tsunami infrequency may imply that the historical record for a given piece of coast is too short to capture a statistical sufficient number of large tsunamis, which entails that tsunami hazard has to be estimated by means of speculated worst-case scenarios, and their consequences are evaluated accordingly and usually result associated with large uncertainty bands. In case of large uncertainties, the main issues for geoscientists are how to communicate the information (prediction and uncertainties) to stakeholders and citizens and how to build and implement together responsive procedures that should be adequate. Usually there is a tradeoff between the cost of the countermeasure (warning and prevention) and its efficacy (i.e. its capability of minimizing the damage). The level of the acceptable tradeoff is an issue pertaining to decision makers and to local threatened communities. This paper, that represents a contribution from the European project TRIDEC on management of emergency crises, discusses the role of geoscientists in providing predictions and the related uncertainties. It is stressed that through academic education geoscientists are formed more to better their understanding of processes and the quantification of uncertainties, but are often unprepared to communicate their results in a way appropriate for society. Filling this gap is crucial for improving the way geoscience and society handle natural hazards and devise proper defense means.

Underwater landslides are hardly predictable and, especially if moving near the coast, represent a concrete threat for the population leaving in coastal areas and for infrastructures placed close to the shoreline. In the framework of the EU funded project TRANSFER, a set of possible sources all around the Mediterranean Sea have been mapped. Among the most common source areas there are the continental margins, owing to their steep slope, favouring the gravitational instability of deposited material. In this work we present the simulation of an ancient event, that occurred during the last glacial age (about 25 kyrs ago), in the Southern Adriatic Sea, known as the Gondola slide. It belongs to the SW Adriatic margin, a margin stretching by about 150 km that is characterized by canyons and widespread failure events that generated slide scars and extensive slide deposits. The most evident slide scar is around 10 km x 2.5 km, at the present-day sea depth of 180 m (but at the time of occurrence the depth was less than 50 meters), with a mobilized volume of around 4.5 km3 and a runout of more than 50 km. Recent bathymetric surveys (high resolution multi beam bathymetry) provided further details on the morphology of the deposit: the upper portion of the slide extends 23 km seaward, down to 800m sea depth, while the distal part is found over 50 km. In this work we consider the tsunami that very likely was generated by such a big slide. The simulation of the slide motion was performed through the code UBO-BLOCK1, developed by the University of Bologna Tsunami Research Team: The model is Lagrangian and block-based, computing the motion and the deformation of each of the blocks in which the total slide mass is discretised. The generation and propagation of the tsunami was simulated through the hydrodynamic code UBO-TSUFE, solving the Navier-Stokes equations in the shallow-water approximation over a mesh formed by triangular elements. The tsunami generation efficiency of Gondola Slide is measured through the Froude number and is found to be low. In spite of the small Froude number value, the tsunami was large because the estimated front of the Gondola Slide is very high with initial mass thickness in the order of several tens of meters. The zone most affected was the coastal region close to the source, that was hit by a long series of waves higher than 8-10 m with period between 15-20 min. The rise of sea level to its present position would lead to reduce the Froude number. Therefore, for a slide like Gondola Slide in identical conditions, but with the today's ocean depth, the associated tsunami would be quite smaller.

We carry out numerical simulations of the tsunami following the Mw=7.4 earthquake which affected the north-western part of Turkey, and in particular the Gulf of Izmit, on August 17, 1999. The earthquake broke at least five long right-lateral strike-slip segments of the North Anatolian Fault, for a total length of about 120 km. The rupture process involved not only strike-slip

TRANSFER (acronym for "Tsunami Risk ANd Strategies For the European Region") is a European Community funded project being coordinated by the University of Bologna (Italy) and involving 29 partners in Europe, Turkey and Israel. The main objectives of the project can be summarised as: 1) improving our understanding of tsunami processes in the Euro-Mediterranean region, 2) contributing to the tsunami hazard, vulnerability and risk assessment, 3) identifying the best strategies for reduction of tsunami risk, 4) focussing on the gaps and needs for the implementation of an efficient tsunami early warning system (TEWS) in the Euro-Mediterranean area, which is a high-priority task in consideration that no tsunami early warning system is today in place in the Euro- Mediterranean countries. This paper briefly outlines the results that were obtained in the first year of life of the project and the activities that are currently carried out and planned for the future. In particular, we will emphasize the efforts made so far in the following directions. 1) The improvement of existing numerical models for tsunami generation, propagation and impact, and the possible development of new ones. Existing numerical models have been already applied to selected benchmark problems. At the same time, the project is making an important effort in the development of standards for inundation maps in Europe. 2) The project Consortium has selected seven test areas in different countries facing the Mediterranean Sea and the eastern Atlantic Ocean, where innovative probabilistic and statistical approaches for tsunami hazard assessment, up-to-date and new methods to compute inundation maps are being and will be applied. For the same test areas, tsunami scenario approaches are being developed, vulnerability and risk assessed, prevention and mitigation measures defined also by the advice of end users that are organised in an End User Group. 3) A final key aspect is represented by the dissemination of the project data and results to the largest possible public. The two privileged means are and will be the project web site (http://www.transferproject.eu) and a web-based GIS database integrating existing data and new project data, ranging from tsunami catalogues to inventories of seismic and non-seismic sources, from topographies and bathymetries at different scales and resolutions to layers containing tsunami inundation maps. Hopefully, this paper will stimulate a discussion and an exchange of experiences with tsunami scientists from different regions of the world.

In the framework of the EU-funded project TRANSFER (Tsunami Risk ANd Strategies For the European Region we faced the problem of assessing quantitatively the tsunami hazard in the Adriatic and north Ionian Seas. Tsunami catalogues indicate that the Ionian Sea coasts has been hit by several large historical tsunamis, some of which of local nature (especially along eastern Sicily, eastern Calabria and the Greek Ionian Islands), while others had trans-basin relevance, like those generated in correspondence with the western Hellenic Trench. In the Adriatic Sea the historical tsunami activity is indeed lower, but not negligible: the most exposed regions on the western side of the basin are Romagna-Marche, Gargano and southern Apulia, while in the eastern side the Dalmatian and Albanian coastlines show the largest tsunami exposure. To quantitatively assess the exposure of the selected coastlines to tsunamis we used a hybrid statistical-deterministic approach, already applied in the recent past to the southern Tyrrhenian and Ionian coasts of Italy. The general idea is to base the tsunami hazard analyses on the computation of the probability of occurrence of tsunamigenic earthquakes, which is appropriate in basins where the number of known historical tsunamis is too scarce to be used in reliable statistical analyses, and the largest part of the tsunamis had tectonic origin. The approach is based on the combination of two steps of different nature. The first step consists in the creation of a single homogeneous earthquake catalogue starting from suitably selected catalogues pertaining to each of the main regions facing the Adriatic and north Ionian basins (Italy, Croatia, Montenegro, Greece). The final catalogue contains 6619 earthquakes with moment magnitude ranging from 4.5 to 8.3 and focal depth lower than 50 km. The limitations in magnitude and depth are based on the assumption that earthquakes of magnitude lower than 4.5 and depth greater than 50 km have no significant tsunamigenic potential. A proper statistical analysis of the catalogue allowed to retrieve the earthquake occurrence rate both at a regional scale and, most importantly, in each of the 30'x30' cells in which the studied geographical domain has been divided into. The final result of the statistical analysis is the computation for each cell of the a- and b-values of a truncated Gutenberg-Richter law. The second step is of more deterministic nature and consists in the tsunamigenic potential determination by using suitable relationships between the earthquake magnitude and the initial disturbance of the sea in each cell. To maximize the coseismic displacement and hence the tsunami initial conditions, only vertical faults have been taken into account. Moreover, each cell has been assigned a typical characteristic focal mechanism (strike-slip or dip-slip) based on the available regional focal mechanism databases and on basic tectonic information. For each magnitude and hence for each initial condition offshore, the tsunami height at the coast is computed through simple empirical amplification formulas. The output of this second step is given by the spatial distribution of the minimum magnitude needed to produce tsunami heights at the coast larger than a given threshold. By combining the results coming from the two steps, we finally determine the number and distribution of tsunamigenic earthquakes expected to occur over a given time interval and to produce tsunami heights larger than a given threshold along any stretch of coast of the selected domain. We will present maps relative to different tsunami height thresholds over a time interval of 10,000 years and discuss the compatibility with the information retrievable from the TRANSFER European tsunami catalogue on one side, and on the other the expected strong relation between the distribution of the parent seismicity and of the resulting tsunami effects, including the importance of doubtful or disputable epicentral determinations for historical earthquakes of moderate to large magnitude.

... TRANSFER, acronym standing for "Tsunami Risk ANd Strategies For the European Region", is a three-year EU-funded research project that tackled all the main fields of interest in tsunami research, ranging from the improvement of the existing tsunami catalogue and the ...