Frameworks for using seasonal climate forecasts for decision making

Farmers in countries such as the Philippines and Australia face a high degree of climate variability. One of the reasons that the climate is variable is because of the El Nino Southern Oscillation (ENSO). Regions that are affected by ENSO tend to have a higher degree of variability, but there is also promise in using ENSO as a slow moving variable which offers some prediction and guidance for the coming season. This guidance from ENSO has been described by Glantz as Science's gift to the 20th Century. In a report to the American Academy of Science, Easterling argued that seasonal climate forecasting based on the interaction of the ocean and atmosphere was one of the premier advances in atmospheric sciences in the 20th century. He went on to say that probabilistic seasonal climate forecasts were ill-suited for decision makers and decision making was ill-suited to probabilistic seasonal climate forecasts. It is clear that when forecasts are presented as a probability it is difficult for users to comprehend what is meant and even harder to know how to use them. It is common for the media and intermediaries to ignore the probabilities and to let a forecast for an El Nino become a forecast for drought rather than an increased chance of drought. There is little difficulty in knowing what to do with a very accurate categorical forecast, as it fits the easy logic of IF, THEN, ELSE. IF the season ahead is going to be a drought, THEN reduce inputs, ELSE continue as normal. The challenge is how to communicate and use in decision making skilful, but uncertain forecasts that are best represented as shifts in climatological probability distributions.It is also common for intermediaries such as agronomists to state that farmers need a categorical forecast because in the end they need to make a decision. Implicit in this statement is the notion that probabilistic forecasts can't be used in decision making.In this paper we report on frameworks whereby information forprobabilistic forecasts can be used in decision making. These are normative frameworks and they may not fit how decision makers currently make decisions. However they do provide useful ways of thinking about uncertainty and distinguishing between lucky and unlucky decisions.