A fuzzy-probabilistic earthquake risk assessment system

Abstract

This paper presents an outline of a risk assessment system for evaluating the expected damage of structures and the consequent financial losses and casualties due to a likely earthquake under elevated uncertain conditions; namely where neither the statistical data nor the seismological and engineering knowledge required for such evaluations are sufficient. In such cases, we should consider extra dimensions of uncertainty, in addition to probability that is usually sufficient for expressing the risk of losses and casualties due to an earthquake where such knowledge and data is available. In the present paper, the uncertainties caused by the insufficient knowledge about the interdependency of various parameters have been considered by means of fuzzy relations. Moreover, the uncertainties in eliciting the likelihood of the seismic hazard have been expressed by fuzzy probability in the form of possibility-probability distributions (PPDs). In other words, fuzzy set theory is employed to complement the standard probability theory with a second dimension of uncertainty. By composition of the fuzzy probability of the seismic hazard and fuzzy vulnerability relation of target structure, the fuzzy probability of damage can be derived. The proposed approach has also been compared with an alternative approach for obtaining a PPD of the hazard. As a case study, the risk assessment system has been tested on a sample structure in the Istanbul metropolitan area.