Decision Analysis in Uncertain Scenarios

Aiding decisions of public interest

Some real world problems are either too complex or occur too rarely to handle easily. Such situations are called “uncertain” in statistics. Professor Simon French's research navigates these complex areas. This work has fed into decision making advice in the event of nuclear accidents for the Scientific Advisory Group for Emergencies (SAGE). The International Atomic Energy Agency (IAEA) also uses Professor French's findings to reduce environmental damage when remediating former nuclear sites.

The challenge

Uncertain scenarios of national and international importance pose two key problems. What is the best approach to decision making under those circumstances? How best can you share this insight with those making decisions at the highest levels? Uncertainty comes in many forms, from inherent uncertainty in a system (such as weather) to incomplete information, due to lack of knowledge, uncertain interrelationships and incomplete data.

With nuclear emergencies being so rare, there is a lack of applicable data. As a result, experts disagree on what would happen should one occur. Nuclear emergencies can produce very different outcomes depending on if the weather is calm or stormy. There are a vast number of spatio-temporal (location and time based) factors to consider. Decision analysis has to interpret the uncertain to allow for authorities to prepare for the most likely scenarios.

Our approach

Professor French’s research followed two main lines of enquiry:

• How can decision analysis fit into political processes? Normally, choices rely on advice from a wide range of stakeholders and experts – who sometimes have opposing points of view on what is “best”.

• Can decision analysis handle uncertainty? Standard theories fail to resolve arguments between experts when data is scarce.

The project used scenario analysis to map possible futures. This allowed Professor French to extend theories of decision making to even the hardest to predict outcomes.

Our impact

Professor French’s research is informing planning at the highest level. This includes the IAEA’s Modelling and Data for Radiological Impact Assessments (MODARIA I and MODARIA II) programmes. Professor French has contributed to two technical documents on existing nuclear site remediation. These to protect public health and the environment alike. Collaboration from 47 IAEA member states showed the international interest in the MODARIA programmes.

Professor French worked with the Met Office and Public Health England on how to prevent the spread of contamination in a radiological emergency. These findings, presented to Scientific Advisory Group for Emergencies (SAGE) of COBR, improved national health protection advice. Professor French has also contributed to the Analysis Under Uncertainty for Decision-Makers Network (AU4DM). The group has collaborated on catalogues to aid decision making with input from industry, government and the third sector.

Learn more about Professor French's other projects

Find out more about the Applied Statistics and Risk Unit (AS&RU)