mainlogo
Tool Catalogue for Uncertainty Assessment
Introduction
1. Sensitivity analysis
2. Error propagation equation ("TIER 1")
3. Monte Carlo analysis ("TIER 2")
4. NUSAP
5. Expert elicitation for uncertainty quantification
6. Scenario analysis
Description
Goals and use
Sorts and locations of uncertainty addressed
Required resources
Strengths and limitations
Guidance on application
Pitfalls
References
7. PRIMA: A framework for perspective-based uncertainty management
8. Checklist for model quality assistance
9. A method for critical review of assumptions in model-based assessments

Pitfalls

Typical pitfalls of scenario analysis are:

  • Undue suggestion of objectivity and completeness: a presentation of e.g. 4 different does not mean that there are only 4 possible ways for the future to develop
  • Quantitative scenarios might suggest numerical exactness and suggest more certainty about the future than we have
  • Analysts and users often forget that outcomes are often based on models which themselves contain assumptions on the future already
  • Scenarios often reflect more our present expectations and beliefs then future developments, and therefore have a tendency to be rather conservative.
  • Scenarios do not forecast what will happen in the future, rather they indicate what might happen under certain well-specified conditions (what-if). Typically scenarios are used in situations where there is lack of information on underlying mechanisms and developments, and therefore it is usually impossible to adhere probabilities to scenarios. Statements on the likelihood of scenarios therefore should be considered with due care.
  • Presenting an uneven number of scenarios may lead users to assume that the middle scenario is the most probable scenario