The Value in Probabilities for Aerodrome Forecasts
Ross Keith
Townsville, Australia
The economic value of a weather forecast is modelled in a signal detection framework as a function of skill, climatological frequency of the event and decision threshold. Derivation of an expression for the optimal decision threshold, which minimises the cost with respect to a perfect forecast, is demonstrated. This optimal decision threshold is a function of the costs of misses and false alarms.
An experiment involving Meteorologists in various offices of the BoM has been in progress for over two years. They have been logging their percentage confidence, at various lead times, that the weather at each airport will be below the alternate minimum. Results are shown by way of relative operating characteristics and reliability diagrams. It is shown that individual forecasters adopt varying decision thresholds, and that the difference in decision thresholds between individuals is more significant than differences in skill.
It is shown that forecasts at lead times less than around 3 hours fail to match persistence, so airlines would be better off using present weather for short domestic flights. An argument is mounted that providing aerodrome forecasts as a probability of breaching the airports alternate minimum would provide a cost outcome superior to the current categorical system. An example for a particular flight will be presented. Furthermore it will be shown that the moderately reliable probability estimates as currently produced by the meteorologists (without any feedback) would, if given in probabilities, produce almost the same savings as perfectly reliable estimates. This would be achieved without any increase in skill.