Gedetailleerde leidraad

A.6 Instructions for filling out the uncertainty matrix

As explained in the foregoing sections, the uncertainty matrix (cf. table 6) employs 5 main dimensions for characterizing the sources of uncertainty: 'location', 'uncertainty level', 'nature of uncertainty', 'qualification of knowledge base' and 'value-ladenness of choices'. These main characteristics have been projected into a two-dimensional matrix:

• by means of the rows one can denote on which specific 'location' the respective uncertainty sources will manifest itself (e.g. a specific uncertain model input, driving force, model parameter etc.).
• by means of the columns one can subsequently indicate how the specific (location-dependent) uncertainty source can be further characterized in terms of the four other qualification-dimensions. Notice that for these dimensions a number of subcategories have been distinguished which enable a differentiated characterization. For the features 'uncertainty level' and 'nature of uncertainty' these sub-characteristics need not be mutually exclusive/disjunct: part of a specific uncertainty source (e.g. a model-input) can be adequately expressed as statistical uncertainty, while another part can e.g. be only expressed in terms of a 'what-if' characterization etc. The last two main columns 'qualification of knowledge base' and 'value-ladenness of choices' provide a reflection on the underpinning and biasedness of the employed knowledge and the choices made, and can usually be characterized in a unique fashion in terms of the corresponding sub-categories (e.g. 'weak/ fair/ strong', resp. 'small/ medium/ large')

It is recommended to use an ABC-coding to indicate the relevance of the item of concern (don't fill in anything if the item is (nearly) not important):

• A= of crucial importance;
• B= important;
• C= of medium importance.

By attaching an index to this coding, e.g. A₁, B₁, C₁, A₂, B₂, C₂ etc., one can indicate to which uncertainty source the respective coding refers: index 1 refers to uncertainty source 1, index 2 to source 2, etc. Notice that a specific source can appear at different points in the matrix with different coding-scores, dependent on how it manifests itself and how it can be characterized.

For reasons of transparency and accountability it is recommended to use the associated table 7 to briefly specify the relevant information on each separate source of uncertainty, and to explain/motivate its uncertainty characterization (location, level, nature etc.) and (A,B,C)-scoring given in the uncertainty matrix. Appropriate literature references or concise background information on these choices can be included..