Evidence-based Reasoning

For model confidence building in particular, a systematic approach has been developed to the organisation of auditable support during model construction together with an innovative approach to the evaluation of uncertainty using evidence-based reasoning. This enables alternative models to be compared on a three-parameter scale comprising 'evidence-for', 'evidence-against' and 'remaining uncertainty' (see below). This evidence-based method can be used to:

• Combine and integrate the use of data and expert judgement in interpretation and modelling,
• Identify explicitly the uncertainties within the system being modelled,
• Explore the significance of particular features or processes of the modelled system through sensitivity analysis,
• Provide comparative analysis between alternative models to allow ranking on the basis of evidence for, evidence against and remaining uncertainty, and,
• Identify the most cost effective areas for targeting data acquisition in order to build confidence / reduce uncertainties in the system to an acceptable level for a decision to be made.

In addition to providing comparative evaluation of detailed geological models arising from detailed site characterisation investigations the evidence-based approach to uncertainty analysis can be used at the preliminary site selection phase, for example to compare the attributes of potential sites based on limited geoscience data but also taking into account preliminary facility design, safety assessment and societal factors.

For further information please contact Mike Egan in the UK or Richard Metcalfe in Japan.