Uncertainty and sensitivity analysis for probabilistic weather and climate risk modelling: an implementation in CLIMADA v.3.1.0

Abstract

Abstract. Modelling the risk of natural hazards for society, ecosystems, and the economy is subject to strong uncertainties, even more so in the context of a changing climate, evolving societies, growing economies, and declining ecosystems. Here we present a new feature of the climate risk modelling platform CLIMADA which allows to carry out global uncertainty and sensitivity analysis. CLIMADA underpins the Economics of Climate Adaptation (ECA) methodology which provides decision makers with a fact-base to understand the impact of weather and climate on their economies, communities, and ecosystems, including appraisal of bespoke adaptation options today and in future. We apply the new feature to an ECA analysis of risk from tropical cyclone storm surge to people in Vietnam to showcase the comprehensive treatment of uncertainty and sensitivity of the model outputs, such as the spatial distribution of risk exceedance probabilities or the benefits of different adaptation options. We argue that broader application of uncertainty and sensitivity analyses will enhance transparency and inter-comparison of studies among climate risk modellers and help focus future research. For decision-makers and other users of climate risk modelling, uncertainty and sensitivity analysis has the potential to lead to better-informed decisions on climate adaptation. Beyond provision of uncertainty quantification, the presented approach does contextualise risk assessment and options appraisal, and might be used to inform the development of story-lines and climate adaptation narratives.

Quantifying Uncertainty and Sensitivity in Climate Risk Assessments: Varying Hazard, Exposure and Vulnerability Modelling Choices

“…Quantitative risk assessment frameworks such as CLIMADA (Aznar-Siguan and Bresch, 2019) provide a useful tool for carrying out global U&SA, as they can be repeatedly applied with minimal computational expense. Kropf et al (2022) present an extension of the CLIMADA platform that seamlessly integrates the SALib -Sensitivity Analysis Library in Python package (Herman and Usher, 2017) into the overall CLIMADA modelling framework, via the UNcertainty and SEnsitity QUAntification (unsequa) module. Following similar steps as a generic U&SA (Pianosi et al, 2016), their approach requires the specification of a probability distribution for each of the uncertain input factors (e.g.…”

“…when the underlying statistical model for the input factor has been derived within the study allowing for a quantification of uncertainty. However, as noted by Kropf et al (2022), when this information is not available it is often not evident how to perturb the inputs, and in particular it is difficult to define physically consistent statistical perturbations of geospatial data.…”

“…We then use the PAWN sensitivity analysis approach of (as implemented in the SAFE toolbox ) to calculate and visualise the sensitivity indices. This approach still relies on expert judgment in terms of defining the discrete set of plausible inputs, and as is common in U&SA, the results will depend greatly on this definition (Pianosi et al, 2016;Kropf et al, 2022). This is discussed further in Section 4.…”

Introducing Uncertainties in the Impact Chain Risk Assessment Framework: The UNTIC tool.

“…To quantify the level of risk imposed by a certain physical disturbance on a natural or socioeconomic system, the three components of risk defined in the ICs framework (i.e., hazard, exposure and vulnerability; Reisinger et al, 2020), are combined in climate risk models (Kropf et al, 2022). In these types of models, indicators for the hazard (i.e.…”

“…However, the different nature of the input data complicates the quantification of the final risk. Particularly, obtaining robust verification data for the exposure and vulnerability components can be quite challenging (Kropf et al, 2022), since these elements are sometimes identified from subjective methods (Zommers et al, 2020).…”