BASE hosts Trust in Models workshop
Scientists from various organisations and disciplines met in Berlin for the specialist workshop Trust in Models at the invitation of the BASE research department on 17 and 18 November. BASE welcomed representatives from the Helmholtz Centre Dresden-Rossendorf, the GFZ Potsdam, the KIT Karlsruhe, the Universities of Clausthal, Stuttgart, Freiberg and Kassel, the BGR, BGE, GRS and NAGRA.
Why is the topic important?
Computer models play a crucial role in many areas of repository safety research and assessment. The interpretation of experimental or geoscientific findings often requires the use of such models. Above all, however, the assessment of repository safety over a million years is not possible without models. The workshop thus focused on the question of what the trustworthiness of models depends on and how it can be improved.
What is the problem?
The answer to this question is multi-faceted. On the one hand, there are the usual standards of quality assurance that models for repository processes are also subjected to. However, such quality assurance cannot verify whether models are appropriate and lead to a correct understanding of the future repository system. To be able to assess this, it is also necessary to examine how models are developed, and what uncertainties exist.
The structure and purpose of models can vary greatly. This means that there is no simple answer regarding the design of trustworthy models. There are simple and complex models, and there are models that are used individually or in combination. Models can represent a realistic forecast or a pessimistic safety estimate. Models also depend on the processes they are supposed to investigate.
Different working groups and fields of application are therefore developing individual perspectives on the topic of modelling.
What was discussed?
The workshop served as an exchange of experiences among scientists, and underlined the complexity of the topic.
It became clear how difficult it is to take an overarching perspective on the complex interaction of all modelling activities. In addition to technical and theoretical aspects, safety-analytical and geoscientific fields of application as well as scientific-theoretical considerations were presented.
What happens next?
The workshop format provided room for discussion, and facilitated an in-depth scientific exchange across disciplinary boundaries. BASE intends to continue this successful dialogue with the scientists in the future, and to expand it to include other topics. A documentation of the current event is in preparation.