Numerical seasonal predictions using global climate models. Researches on the long range forecasting at Météo France

Abstract.

 Since 1995, Météo France has engaged important research works concerning seasonal forecasting within the framework of projects of the European Union. One of these projects is described here. The main goal of the project PROVOST was to evaluate the potential of predictability on seasonal and monthly time-scales of some Global Climate Models (GCM) running separately ensembles of integrations, or gathered in a multimodel ensemble of predictions. The result is that the deterministic skill and the probabilistic skill are both improved by increasing the ensemble size and by merging different models.