Management, visualization and comparison of multiple hazards and risk using free software: the ByMuR tool

Abstract

Nowadays, risk assessment is an essential tool for risk management and risk mitigation policies since it provides an evaluation of the risk associated to a recognized threat that can cause harm or damage to humans, their structures or the environment. Risk results from the combination of three factors (UNDRO, 1982): Risk = Hazard x Exposure x Vulnerability. In this very general framework, the hazard is a measure of the likelihood of the expected size of a dangerous phenomenon in a given target area and in a given exposure time, the exposure measures the assets that are potentially at risk (both people and structures) and the vulnerability represents the degree at which the assets may be damaged due to the hazard. Of course, this represents a simplification of the reality, which does not account for any uncertainty. Modern risk assessment is based on Probabilistic Risk Analyses (PRA), making use of probabilistic hazard analysis, fragility and loss models. Such probabilistic factors are typically combined the through Pacific Earthquake Engineering Research (PEER) formula [Cornell and Krawinkle, 2000; Der Kiureghian, 2005], which has been originally proposed for seismic risk and it has been recently proposed as a possible base for multi-risk assessments [Marzocchi et al., 2012; Selva, 2013; Mignan et al., 2014; Liu et al., 2015]. In this framework, the project called ByMuR (http://bymur.bo.ingv.it), funded by the Italian Ministry of Education, Universities and Research (Ministero dell’Istruzione, dell’Università e della Ricerca – MIUR) and lasted 4 years from November 2010 to November 2014, has been accomplished with the goal of providing a quantitative and objective method for quantifying the multi-risk for a given area. The approach is based on Bayesian inference that allows to account for both aleatory and epistemic uncertainties along all the phases of the computation, from hazard to risk (for hazards: [Selva and Sandri, 2013; Faenza et al., 2017; Sandri et al., 2016; Grezio et al., 2015; Selva et al., 2015]; for vulnerability and risk: [Selva et al., 2013; Selva et al., in prep.]; for the uncertainty treatment: [Marzocchi et al., 2015]). More specifically, the project was focused to: (i) provide a quantitative and objective general method for a comprehensive long-term multirisk analysis in a given area, accounting for inter-model epistemic uncertainty through Bayesian methodologies, and (ii) apply the methodology to seismic, volcanic and tsunami risks in Naples (Italy).
Figure 1. General scheme of all the inputs to ByMuR software for the case study of Naples. The tool handles both single and multi risk factors through a database and users can visualize or combine them to perform risk/multi-risk analysis for a selected target area.
One of the main products/results provided by the project is the homonym ByMuR software [Perfetti and Tonini, 2018], an open source tool aiming to manage, visualize and compare PRA results (e.g. risk curves, risk indexes) for different hazards, as well as, all the components of any PRA (e.g. hazard, fragilities and exposure), within a multi-risk perspective. For each component and for PRA results, epistemic uncertainty is handled through the mean and several percentiles of the community distribution (e.g., SSHAC 1997) or of the ensemble model [Marzocchi and Jordan, 2014]. The ByMuR software handles separately the probabilistic hazard assessments of different kind of hazardous phenomena, the relative fragility and loss models, exposure data, as well as the corresponding probabilistic risk results. The software manages precomputed results, which are input through standard formats and allows several post-processing of such data in order to produce ensemble hazard models and spatially aggregated results. In this report we present the implementation of the ByMuR software, by describing its features and by illustrating how it could serve the scope of supporting a multi-hazard and multi-risk analysis, as well as, for any hazard assessment. The results computed for the municipality of Naples are demonstrative of the general methodology and, at the same time, are here used to illustrate the features and usability of the ByMuR software. The case study is focused on the long-term (5, 10 and 50 years) multi-risk assessment due to volcanic (tephra fall and pyroclastic flows), seismic and tsunami hazardous phenomena.