Empirical building fragilities from observed damage in the 2009 South Pacific tsunami

This manuscript presents empirical building fragility functions that were developed based on data obtained from the 29th September 2009 South Pacific tsunami. A multi-disciplinary reconnaissance team involving topographic surveyors, tsunami/hydrology modellers, structural engineers, and risk analysts collected observational and quantitative data on building damage and the tsunami demand. A diverse range of collected data, which included a topographic survey, observed water depths, predominant flow directions, inundation limits, building damage, and eyewitness reports, were used to recreate the peak tsunami-induced demands on structures at each of the survey locations. Using the interpreted data, fragility functions were developed for a variety of building classes using logistic regression. It was observed that residential timber structures were more fragile (i.e. have a higher likelihood of damage for a given water depth) than masonry residential structures for severe and collapse damage states. Conversely, residential reinforced concrete structures were observed to be less fragile than residential masonry structures for severe and collapse damage states. The influence of 'shielding' and 'entrained debris' effects on fragility functions were also quantified using the empirical data. Other parameters such as velocity or impact duration were not considered due to the paucity of data. The results of this study contribute to the ongoing development of robust methods for explicitly estimating tsunami-induced risk to coastal communities.