Author(s): Alessandro Pagano, Raffaele Giordano, Irene Pluchinotta, Anna Bruna Petrangeli, Umberto Fratino, Michele Vurro
Linked Author(s): Umberto Fratino
Keywords: Bayesian Belief Networks (BBNs), vulnerability assessment, water supply systems, uncertainty management, Disaster Risk Reduction (DRR)
Abstract: Drinking water availability and quality is crucial for the well-being and the safety of people, particularly in case a disaster occurs. In fact, the provision of critical services in general, particularly of drinking water after an emergency, may significantly contribute to limit the intensity and the duration of crises and support the resilience of the whole community. Among the main issues that contribute to make emergency management and decision-making into complex processes, it is worth to underline the role of uncertainty, which mainly representing the lack of reliable data and information related to the impacts of a disaster after its occurrence. Particularly focusing on drinking water supply systems, the development of tools capable of operating in a framework of uncertainty, quickly identifying the spatial location of the most critical elements of a complex infrastructure, assessing risk levels and suggesting potential strategies to cope with a disaster, may be crucial for decision-makers. The present article describes the implementation, using the potentialities of GIS systems, of a vulnerability assessment model for drinking water supply infrastructures based on Bayesian Belief Networks (BBNs). The tool is implemented in L�Aquila case study, which is particularly relevant in the recent history of emergencies related to natural hazards and contributes to stress the key role of either modeling or information of the uncertainty in the emergency management. BBNs are capable to integrate and manage different classes of data and categories of information, and to deal with information having a different level of uncertainty. Uncertainty, which is integral to Bayesian analysis, can be used to provide decision-makers with reliable information to select the most suitable strategies to manage emergencies. In the present work, a quantitative approach to uncertainty modeling is provided as well, and coupling with modeling results should support the effectiveness of decision-making processes
Year: 2017