Publication
As our climate continues to respond to anthropogenic forcing, the magnitude and frequency of individual weather events and the intensity of the weather extremes associated with these, re -mains highly uncertain. This is a particular concern for our infrastructure networks, as increasing storm-related damage to these vital lifelines has significant consequences for our communities. Effective first response is hence becoming an increasingly important part of the management of infrastructure systems. Here, we propose a novel and rational framework for 'consequence forecasting' that enables probabilistic, pre-event decision-making for first responders to effec-tively target resources prior to an extreme weather event and thus reduce the societal conse-quences. Our method is unique in that it minimises model bias by using the same numerical weather prediction model for both fault attribution and fault prediction. Our framework can predict failure rates that are within 50% of the true value for more than 50% of the events considered, some 24 h in advance, therefore demonstrating that it can be an important part of increasing societal climate resilience by reducing reinstatement times.