Seismic risk analysis and management of civil infrastructure systems: an overview - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 139

theory and Italian fl ags to earthquake engineering risk management prob-
lems is presented in Chapter 9.
The role of quantitative/qualitative risk analysis is to provide objective
evidence on potential risk by utilizing available scientifi c tools. Owing to
unavoidable uncertainties and complexities of civil engineering structures
under seismic risk, risk assessment and management are infl uenced by
various factors (beyond physical characteristics of engineering materials
and systems). They include risk perception (e.g. expert versus public), risk
criteria (individual versus societal, or voluntary versus involuntary), politi-
cal process, and risk communication. Depending on history of past events
and public concern/reaction, objective risk assessment results may be dis-
puted and on occasion overruled. Nevertheless, it is important to develop
quantitative decision support tools for earthquake risk mitigation. Such
tools are useful to quantify/compare seismic risks for different options and
to facilitate informed decision making. Aiming at summarizing salient fea-
tures of available analytical and decision making frameworks for assessing
seismic risk quantitatively, several key seismic risk analysis tools are dis-
cussed in Sections 6.3 and 6.4. Specifi cally, Section 6.3 is focused on the
acceptable risk criteria for different stakeholders (e.g. F - N curves and
life quality index). In Section 6.4, measures for seismic risk management,
such as earthquake insurance, retrofi tting, and strengthening, are discussed
from cost-benefi t analysis perspective as well as multiple criteria decision
making perspective, through illustrative examples. Subsequently, conclu-
sions and future trends of seismic risk analysis and management are
mentioned.

6.2 Uncertainty in risk analysis

Traditionally, aleatory and epistemic uncertainties are referred to as risk
and uncertainty, respectively (Paté-Cornell 1996). Knight (1921) made a
distinction between risk and uncertainty; risk is something to which math-
ematical probabilities can be assigned either through a priori knowledge or
from the statistics of past experience, whereas uncertainty is referred to as
randomness, which cannot be explained. Blockley (1995) and Klir and Yuan
(1995) have broadly categorized epistemic uncertainty into vagueness and
ambiguity . Ambiguity is due to unclear distinction of various alternatives,
which is further divided into discord (confl ict) and non-specifi city. Vague-
ness (imprecision) refers to lack of defi nite or sharp distinction. The tax-
onomy of uncertainty, albeit to a different degree, is refl ected in seismic risk
analysis. Indeed, ignorance and variability require different methods of
uncertainty propagation (Ferson and Ginzburg 1996). Subsequently, Klir
(2004) proposed a 'generalized information theory' (GIT) to develop
broader treatment of uncertainty quantifi cation.