What can we learn about our forecast trigger from just a single activation?

Through a mock debate, this session explored different perspectives on four key questions that stakeholders often discuss after an activation:

• Should we change the forecast model if a different model forecasted the event better?
• Should we change the probability threshold if we acted in vain or did not trigger?
• Should we change the impact level if the impacts were different to those forecast?
• Should we change things if we are triggering too frequently?

The views presented during the debate as well as the ensuing discussion made clear that there are no easy answers. As summarised by one audience member “how do we determine what level of trade-off is acceptable given the expectations of members, stakeholders and affected communities?”
What looks like success from the perspective of a modeller may be viewed very differently by a humanitarian or someone affected by a crisis. While it is impossible to reach a single conclusion on such complicated topics, the dialogue underscored the importance of ongoing conversations about models and balancing the potential for acting in vain with humanitarian impacts in each situation.