HYRISK: An R package for hybrid uncertainty analysis using probability, imprecise probability and possibility distributions

Abstract

Uncertainty analysis is an unavoidable risk assessment task (for instance for natural hazards, or for environmental issues). In situations where data are scarce, incomplete or imprecise, the systematic and only use of probabilities can be debatable. Over the last years, several alternative mathematical representation methods have been developed to handle in a more flexible manner the lack of knowledge related to input parameters of risk assessment models. This article presents an R package HYRISK dedicated to jointly handling different mathematical representation tools, namely probabilities, possibility distributions and probability functions with imprecise parameters, for the different stages of uncertainty treatment in risk assessments (i.e. uncertainty representation, propagation, sensitivity analysis and decision-making). We support the description using the case study of a dike stability analysis. The package is available at: https://cran.r-project.org/web/packages/HYRISK/index.html.

Deep uncertainties in shoreline change projections: an extra-probabilistic approach applied to sandy beaches

“…Epistemic uncertainty is represented by the breadth between the upper and lower CDFs, whereas aleatory uncertainty is represented by the overall tilt of the p box. The gap between the upper and lower CDFs can be considered "what is unknown" and represents the imperfect state of knowledge (Rohmer et al, 2019). To quantify this deep uncertainty, we use an indicator termed "ambiguity" and defined as the width (in metres) between medians of the upper and lower CDFs.…”

“…6 reveal that the uncertainty strongly amplifies with distant time horizons, in particular under high-global-warming scenarios. From a coastal planning perspective, such large uncertainties can be considered unhelpful and not be used as such to support the decision-making process (Rohmer et al, 2019). In this case, it is particularly relevant to determine which uncertainty contributes the most to the total uncertainty in order to anticipate how foreseen improvements in the understanding of the physical system could reduce the uncertainty of projections.…”

“…In contrast, when uncertainty tolerance is low, robust decision-making is preferable, which implies testing adaptation options against any plausible scenarios, hence considering high-end (Hinkel et al, 2015(Hinkel et al, , 2019Kopp et al, 2017;Stammer et al, 2019) and low-end (Le Cozannet et al, 2019b) projections (or scenarios), which explore plausible -although unlikely -upperand lower-tail sea level scenarios beyond the likely range. Although the literature above has focused on sea level rise information needs, the same type of information is needed for its coastal impacts (Rohmer et al, 2019), raising the need for a framework allowing the propagation and analysis of deep uncertainties from sea level rise to its impacts.…”

“…To date, both types of uncertainties have been addressed mainly using the tools provided by the probability theory and occasionally used in combination with expert knowledge (especially for sea level projections; Oppenheimer et al, 2019;Bamber et al, 2019). Yet, several studies have pointed out that the use of probabilities merges the different uncertainty types in a single format and can in turn induce an appearance of overconfidence in uncertainty quantification (Le Cozannet et al, 2017;van der Pol and Hinkel, 2019;Bakker et al, 2017;Rohmer et al, 2019). Such a misleading effect can have a serious impact on coastal risk management and planning.…”

“…These are termed extra-probabilistic because they avoid the selection of a single probability law by bounding all the possible probability models consistent with the available data. The added value of these approaches has been discussed for global SLR projections (Le Cozannet et al, 2017) or to assess local flood impact (Rohmer et al, 2019) but has never been used in the context of coastal erosion to our knowledge.…”