Flood Forecast Verification to Support Emergency Management

The aim of the paper is to set out a method with which to evaluate Flood Early Warning Systems (FEWSs) in regard to their usefulness in warning decision-makers. In practice, this equates to carrying out flood forecast verification in order to support the warning process or, more generally, the emergency management. A procedure is proposed for conduct of the evaluation in different warning contexts. Its application to a case study is also described. A critical analysis highlights that the procedure is flexible and comprehensive, so that it can be implemented in regard to different decisionmaking problems. The main innovation consists in the inclusion of damage among the variables on which FEWSs performance assessment is grounded. In spite of marked uncertainties affecting the complex modelling chain, the results give significant indications about the behaviour of the system.