Expert judgement is routinely required to inform critically important decisions. While expert judgement can be remarkably useful when data are absent, it can be easily influenced by contextual biases which can lead to poor judgements and subsequently poor decisions. Structured elicitation protocols aim to: (1) guard against biases and provide better (aggregated) judgements, and (2) subject expert judgements to the same level of scrutiny as is expected for empirical data. The latter ensures that if judgements are to be used as data, they are subject to the scientific principles of review, critical appraisal, and repeatability. Objectively evaluating the quality of expert data and validating expert judgements are other essential elements. Considerable research suggests that the performance of experts should be evaluated by scoring experts on questions related to the elicitation questions, whose answers are known a priori. Experts who can provide accurate, well-calibrated and informative judgements should receive more weight in a final aggregation of judgements. This is referred to as performance-weighting in the mathematical aggregation of multiple judgements. The weights depend on the chosen measures of performance. We are yet to understand the best methods to aggregate judgements, how well such aggregations perform out of sample, or the costs involved, as well as the benefits of the various approaches. In this paper we propose and explore a new measure of experts’ calibration. A sizeable data set containing predictions for outcomes of geopolitical events is used to investigate the properties of this calibration measure when compared to other, well established measures.

Our main result gives the asymptotic distribution of the determinant of a random correlation matrix sampled in a particular way from the space of d×d correlation matrices. Several spin-off results are proven along the way, and an interesting connection with the law of the determinant of general random matrices is investigated. As different methods for generating random correlation matrices are proposed in the literature, one application of our result is that in can be employed to differentiate between those methods.

Probabilistic thinking can often be unintuitive. This is the case even for simple problems, let alone the more complex ones arising in climate modelling, where disparate information sources need to be combined. The physical models, the natural variability of systems, the measurement errors and their dependence upon the observational period length should be modelled together in order to understand the intricacies of the underlying processes. We use Bayesian networks (BNs) to connect all the above-mentioned pieces in a climate trend uncertainty quantification framework. Inference in such models allows us to observe some seemingly nonsensical outcomes. We argue that they must be pondered rather than discarded until we understand how they arise. We would like to stress that the main focus of this paper is the use of BNs in complex probabilistic settings rather than the application itself.

Emergency management and long-term planning in coastal areas depend on detailed assessments (meter scale) of flood and erosion risks. Typically, models of the risk chain are fragmented into smaller parts, because the physical processes involved are very complex and consequences can be diverse. We developed a Bayesian network (BN) approach to integrate the separate models. An important contribution is the learning algorithm for the BN. As input data, we used hindcast and synthetic extreme event scenarios, information on land use and vulnerability relationships (e.g., depth-damage curves). As part of the RISC-KIT (Resilience-Increasing Strategies for Coasts toolKIT) project, we successfully tested the approach and algorithm in a range of morphological settings. We also showed that it is possible to include hazards from different origins, such as marine and riverine sources. In this article, we describe the application to the town of Wells-next-the-Sea, Norfolk, UK, which is vulnerable to storm surges. For any storm input scenario, the BN estimated the percentage of affected receptors in different zones of the site by predicting their hazards and damages. As receptor types, we considered people, residential and commercial properties, and a saltmarsh ecosystem. Additionally, the BN displays the outcome of different disaster risk reduction (DRR) measures. Because the model integrates the entire risk chain with DRR measures and predicts in real-time, it is useful for decision support in risk management of coastal areas.