Assessing time, cost and quality trade-offs in forecast-based action for floods

Forecast-based actions are increasingly receiving attention in flood risk management. However, uncertainties and constraints in forecast skill highlight the need to carefully assess the costs and benefits of the actions in relation to the limitations of forecast information. Forecast skill decreases with increasing lead time, and therefore, an inherent trade-off between timeliness and accuracy exists. In our paper, we present a methodology to assess the potential added value of early warning early action systems (EWEAS), and we explore the decision-makers’ dilemma between acting upon limited-quality forecast information and taking less effective actions. The assessment is carried out for one- and a two-stage action systems, in which a first action that is based on a lower skill and longer lead time forecast may be followed up by a second action that is based on a short-term, higher-confidence forecast. Through an idealized case study, we demonstrate that a) that the optimal lead time to trigger action is a function of forecast quality, the local geographic conditions and the operational characteristics of the forecast-based actions and b) low-certainty, long lead time forecasts can become valuable when paired with short-term, higher quality ones in a two-stage action approach.