Answers are grounded in this paper and cite their sources.
In the face of possible extreme weather, such as a tropical cyclone landfall, forecast probabilities can help with the decision of whether to evacuate or stay. Such decisions can be studied in an idealized sense using the well-known cost-loss model. A more subtle but perhaps more relevant question that also arises in many evacuation situations is whether to evacuate now or wait for the next forecast. We study this question, again in an idealized sense, using a new cost-loss model. We investigate what the model implies in terms of the weather forecast information required to make optimal decisions. We find that uncertainty around the probability of extreme weather does not make any difference to decisions. However, we find that information about how the probability might change between forecasts does make a difference. Based on this result, we then consider probability forecasts presented as a probability and a standard deviation of change. We give an example in which low standard deviations imply evacuation as the best option, while higher standard deviations imply waiting for the next forecast as the best option. Idealized models of this type are not intended to be used as actual decision making algorithms in real situations. However, they can help us understand the logic of the decision-making process, and may lead to an improved understanding of what information could be provided that could help decision making. Our evacuation cost-loss model demonstrates that evacuation decisions could potentially be improved if forecast probabilities included a standard deviation of change.
No takes yet. Share an insight, caveat, or question.
Stephen Jewson (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: