The implications of empirical data for risk

It is of interest for national governments to assess strategic issues such as natural hazards and anthropogenic threats with some reference to risk, in order to support prioritisation of treatment solutions. With most threats of strategic relevance such as earthquakes, pandemics and terrorism following a distribution in size of events, representation of the risk for a threat as a single frequency-consequence pair is often inadequate as this single pair may exclude a significant portion of data and their contribution to total risk. Identifying the entire distribution of event sizes and their frequencies is better suited for understanding the relative contributions to total risk from high and low consequence events. If the distribution of event sizes does follow a law, the finite size of data-sets makes identifying the law difficult. This paper outlines the steps required to utilise empirical data to inform the risk of strategic threats and support decision-makers to prioritise treatment options according to their relative contributions to total risk. Potential pitfalls and limitations are also described.