Comparing COSMO-LEPS and COSMO-SREPS in the short-range.

In this paper, the impact of the two different perturbation strategies for boundaries on shortrange forecast skill is assessed. This work is part of the COSMO Priority Project CONSENS, which aims at deciding which strategy is better to follow for short-range ensemble forecasting at the mesoscale. The impact of combining the two ensembles in different ways, for the aim of short-range forecasting, is also analysed. Results show that there is a positive effect in adding multi-model-driven members to the COSMO-LEPS ensemble, thus providing a hybrid solution to the problem.