Tsunami are rare geological phenomena but their impacts are frequently catastrophic. Greece is the European country that has been affected most frequently by tsunami. Whilst tsunami hazard and risk within Greece have been studied, tsunami vulnerability has not. Previous studies and hazard mapping of tsunami flood prone areas around the world imply that populations, buildings and infrastructure within a potential inundation zone are uniformly vulnerable to impact. However, vulnerability is a dynamic factor and is a function of a number of parameters, since vulnerability of any physical, structural or socio-economic element is the probability of it being damaged, destroyed or lost. This dissertation identifies two coastal segments within Greece known to be at risk from tsunami flooding; reviews the historical record of events and determines a worse case tsunami event for both locations. A new methodology for assessing vulnerability of these coastal communities is then developed and applied to each location. The methodology incorporates multiple parameters that contribute to tsunami vulnerability and identifies and demonstrates the vulnerability of different components of the community (human, economic and the built environment). Using a Multi Criteria Evaluation Method the vulnerability of individual buildings and open spaces within the potential inundation zone are assessed and the results are displayed in map form with the aid of a Geographical Information System (GIS). The primary database may be used by various end-users, such as emergency planners, insurance companies, individuals, local authorities and scientists, in order to produce maps that could enable them to make recommendations regarding response plans or mitigation measures. Finally, a framework for Tsunami Vulnerability Assessment is introduced which integrates this vulnerability assessment methodology. Through this framework, the Greek Civil Protection system is outlined, its weaknesses are discussed and recommendations are made.

One of the most catastrophic natural hazards which can devastate coastal zone communities is the tsunami. The risk of tsunami devastation can be mitigated by reconstruction and quantification of past tsunamis, but this requires identification and analysis of past tsunami magnitudes and dates even from historical times. The interpretation of geological records of tsunamigenic deposits is the core theme explored in this thesis for the characterisation of palaeotsunamis. Three key elements: identification of palaeotsunamites, dating their deposition and determination of magnitude are needed for reconstruction and quantification of palaeotsunamis. While several studies have described criteria for identification of tsunamigenic sediments, much less research has been carried out on reconstruction of palaeotsunamis. Sediment characteristics, their depositional configuration and extent of the inundation area on the coast give infOlmation on hydrodynamic conditions of tsunami waves. This study has developed a method to estimate tsunami risk by reconstructing and quantifying palaeotsunamis from tsunamigenic sediments on the Sri Lankan coast rising sedimentological and paleontological characteristics, Ground Penetrating Radar (GPR), and optically stimulated luminescence (OSL) dating techniques. The 2004 tsunami records were used as a control for the study. Distinctive tsunamigenic sediment signatures were recognised. Three palaeotsunami events were identified and dated at ISO (Krakatua), 2550±190 and 3170±320 years BP and correlated with historical and archaeological records. The recurrence interval of c 600 years postulated for the Indian Ocean was confirmed by this study and the oldest event recorded to date in the Indian Ocean was recognised at 3170±320BP. The method developed here based on reconstruction of palaeotsunamis using sedimentological records, GPR and OSL techniques enables estimation of inundation distance, recurrence interval and consequently the wave characteristics of palaeotsunamis which can be used to estimate tsunami risk for any coast.