A Bayesian Hierarchical Approach to Ensemble Weather Forecasting

In meteorology, the traditional approach to forecasting em- ploys deterministic models mimicking atmospheric dynamics. Forecast uncertainty due to the partial knowledge of initial conditions is tackled by Ensemble Predictions Systems (EPS). Probabilistic forecasting is a relatively new approach which may properly account for all sources of uncertainty. In this work we propose a hierarchical Bayesian model which devel- ops this idea and makes it possible to deal with an EPS with non-identifiable members using a suitable definition of the sec- ond level of the model. An application to Italian small-scale temperature data is shown.