Robust Decisions Under Uncertainty: examples of info-gap analysis in mitigation policy and flood risk management
Robustness has been widely identified as a desirable attribute of decisions that have to be taken under conditions of severe uncertainty.
We think of a decision being robust to uncertainty if it yields acceptable outcomes under a wide range of possible futures. Yet robustness is seldom cost free. It usually entails sacrificing some performance relative to an ‘optimal’ solution. Uptake of robust solutions therefore needs to be justified by methodology that can be used to compare options on the basis of robustness and the cost at which it is achieved.
One such approach is info-gap theory, which was originally developed by Yakov Ben-Haim, and has now seen application in a very wide range of fields, including ecology, economics and engineering.
This talk provided a brief introduction to info-gap theory before presenting examples of its application to mitigation and adaptation decisions.
The mitigation example used a version of the DICE model that includes the (uncertain) probability of Atlantic Meridional Overturning Circulation collapse and its consequences.
The adaptation examples dealt with flood defences on rivers and in the Thames Estuary.