Citation

Riga E, Karatzetzou A, Mara A, Pitilakis K. Soil Dyn. Earthqu. Eng. 2017; 92: 9-24.

Copyright

(Copyright © 2017, Elsevier Publishing)

DOI

10.1016/j.soildyn.2016.09.043

PMID

unavailable

Abstract

Seismic risk assessment and loss estimation are of major importance for decision-making with respect to the reduction of earthquake-induced losses in large urban areas. However, the methodological chain of seismic risk assessment, from seismic hazard assessment to evaluation of potential losses, encompasses numerous uncertainties, both aleatory and epistemic, associated with different sources. The present study is a comprehensive application of the Capacity Spectrum Method to the seismic risk assessment of the city of Thessaloniki, aiming to give an insight into epistemic uncertainties involved in the above methodology, owing to hazard modelling, structural capacity, fragility and damping, as well as shaking duration. To quantify and discuss the uncertainties, a logic tree approach is used. A sensitivity analysis of the computed seismic risk results is performed to determine the input parameters having the greatest impact. The analyses were carried out for the building stock of the city of Thessaloniki, Greece, for which detailed building inventory and very good knowledge of the soil conditions are available. Only physical losses due to the structural damage of the building stock were considered. Considerable scatter in the risk estimates was observed due to epistemic uncertainties. The sensitivity analyses demonstrated that the most influencing parameters when applying the Capacity Spectrum Method are the selection of the fragility curves for the buildings and the seismic hazard model adopted in the analysis. The decision-making process with respect to seismic risk assessment should therefore carefully account for uncertainties and pay attention to the most influencing parameters regardless of the methodology used.

Language: en

Keywords

Capacity spectrum method; Logic tree; Seismic risk assessment; Sensitivity analysis; Uncertainties