Multi-hazard coupling disasters, in which multiple hazards occur simultaneously and interact to compound the consequences, are a common phenomenon. The assessment of the individual risk in multi-hazard coupling disasters faces several difficulties due to the nonlinear additivity of risks from multiple hazards. This article presents the Choquet integral multiple linear regression model as a method of overcoming the problems of nonlinear additivity. Using this method, the nonlinear additive individual risks of multi-hazard coupling disasters can be superposed with the nonadditivity of the fuzzy measure during the Choquet integral and the nonlinearity of the Choquet integral itself. This method also takes into account the effects of magnification on the severity of disasters and the vulnerability of victims in multi-hazard disasters. It provides the magnification coefficients to quantitatively calculate the risks of all disasters. To examine the efficacy of the risk-assessment measure, this article uses as a case study the severe fire and explosion disaster that occurred in a port at Tianjin, China, in 2015. From this case study, it can be concluded that the composite individual risk of multi-hazard coupling disasters is greater than that of the simple addition of the risk of each hazard. This finding indicates that multi-hazard coupling disasters are more severe than disasters involving single hazards. Moreover, this risk-assessment method provides guidance in preventing, estimating, and dealing with multi-hazard coupling disasters. It can also provide solutions to complex risk-analysis problems in fields, such as finance, economics, and information science.
Keywords: Choquet integral; fuzzy measure; magnification effect; multi-hazard coupling disaster.
© 2020 Society for Risk Analysis.