Decision theory and the analysis of rare event space weather forecasts

摘要: [1] Several basic results from decision theory as applied to rare event forecasts are reviewed, and an alternative method for comparing is presented. A fundamental result that a large class of users only interested in economic utility, the relevant performance quantity number correct false alarm forecasts. This contrasted with reality most forecast models optimized have high data-model correlation, which does not always correspond maximum utility. The value score (VS) developed by Wilks (2001) partially resolves this disconnect between modeler- user-relevant metrics. Although closer what likely interest user, maximal VS necessarily utility realistic case where cost benefit dependent on amplitude forecasted event. An comparison presentation proposed may resolve problem. For considered, full specification model requires computation probability correct, alarm, missed at several levels warning time spans. Examples computations involved modeler user given predictions large-amplitude energetic electron fluence geomagnetic storms parameterized Dst index.