The thesis has for objective to study new statistical methods to correct temperature predictionsthat may be implemented on the ensemble prediction system (EPS) of Meteo France so toimprove its use for the electric system management, at EDF France. The EPS of Meteo Francewe are working on contains 51 members (forecasts by time-step) and gives the temperaturepredictions for 14 days. The thesis contains three parts: in the first one we present the EPSand we implement two statistical methods improving the accuracy or the spread of the EPS andwe introduce criteria for comparing results. In the second part we introduce the extreme valuetheory and the mixture models we use to combine the model we build in the first part withmodels for fitting the distributions tails. In the third part we introduce the quantile regressionas another way of studying the tails of the distribution.
Meteorological and climate data are indeed essential both in day-to-day energy management and for the definition of production and distribution infrastructures. For instance, the supply of electricity to users can be disturbed by extreme meteorological events such as thunderstorms with unusually strong winds, severe icing, severe cold spells, sea level elevation associated with storm surges, floods ... To be protected against such events, it is not sufficient to act after they have taken place. It is necessary to identify their potential impacts precisely and assess the probability of their occurrence. This book shows that this can only be done through an enhanced dialogue between the energy community and the climate and meteorology community. This implies an in-depth dialogue between actors to define precisely what kind of data is needed and how it should be used. Météo-France has been in long-term cooperation with the energy sector, including the fields of electricity production and distribution. Drawing on this experience, it should be noted in this respect the importance of lo- term partnership between actors as exemplified here by the message of EDF.