Earthquake risk assessment and management : case study, Cyprus
Earthquakes are amongst the worst natural disasters on Earth, resulting in an annual average of around 10,000 fatalities last century and progressively increasing in the amount of economic damage they cause, reaching US $20 billion per annum this decade. The mitigation of the unwanted consequences of earthquakes is normally achieved by Risk Management Strategies (RMS), which rely on the development of Earthquake Risk Assessment (ERA) techniques. This thesis aims to develop a framework for ERA for medium seismicity regions that incorporates the spatial aspects of the hazard and risk evaluation. The framework is used to undertake ERA for the island of Cyprus, and the information is used to propose RMS. The ERA framework relies on comprehensive data on the location, value and vulnerability of buildings and the population distribution. These data were collected from the various Cyprus Government Departments. Various hazard and attenuation models are examined, and the effect of their variability is taken into account through Monte Carlo simulations. The estimated annual risk for Cyprus is just below £ 10 million CY. This value was estimated based on the use of the re-appraised historical data for the past-century. Comparisons with other seismic hazard assessment methods, such as recurrence relationships, have revealed that, without a spatial distribution model, such approaches are unsuitable for ERA. Though the maximum intensities predicted are in line with the ones that underpin the aseismic code of Cyprus (CCEAA-CFEE, 1994), the predicted design accelerations are higher than given in the code. Hence, new seismic accelerations are proposed. Despite that, the current reduction in risk is comparable to the additional cost of aseismic design. Seismic retrofitting was also examined and it was found that as part of a general modernisation scheme seismic upgrading is cost effective. However, whatever the state of the building, it is recommended that earthquake insurance should be made mandatory. The current seismic insurance rates appear to be fair, though they seem to underestimate the risk in the areas of high seismicity. The number of likely human losses is also estimated. This study concludes that the result of ERA is heavily dependent on the models and data used, and both require constant updating for the ERA results to remain meaningful.