caa a22 4500 605473609 CHVBK 20210128100343.0 cr unu---uuuuu 210128e20150401xx s 000 0 eng 10.1007/s00382-014-2344-9 doi (NATIONALLICENCE)springer-10.1007/s00382-014-2344-9 Impact of the initialisation on the predictability of the Southern Ocean sea ice at interannual to multi-decadal timescales [Elektronische Daten] [Violette Zunz, Hugues Goosse, Svetlana Dubinkina] In this study, we assess systematically the impact of different initialisation procedures on the predictability of the sea ice in the Southern Ocean. These initialisation strategies are based on three data assimilation methods: the nudging, the particle filter with sequential importance resampling and the nudging proposal particle filter. An Earth system model of intermediate complexity is used to perform hindcast simulations in a perfect model approach. The predictability of the Antarctic sea ice at interannual to multi-decadal timescales is estimated through two aspects: the spread of the hindcast ensemble, indicating the uncertainty of the ensemble, and the correlation between the ensemble mean and the pseudo-observations, used to assess the accuracy of the prediction. Our results show that at decadal timescales more sophisticated data assimilation methods as well as denser pseudo-observations used to initialise the hindcasts decrease the spread of the ensemble. However, our experiments did not clearly demonstrate that one of the initialisation methods systematically provides with a more accurate prediction of the sea ice in the Southern Ocean than the others. Overall, the predictability at interannual timescales is limited to 3years ahead at most. At multi-decadal timescales, the trends in sea ice extent computed over the time period just after the initialisation are clearly better correlated between the hindcasts and the pseudo-observations if the initialisation takes into account the pseudo-observations. The correlation reaches values larger than 0.5 in winter. This high correlation has likely its origin in the slow evolution of the ocean ensured by its strong thermal inertia, showing the importance of the quality of the initialisation below the sea ice. Springer-Verlag Berlin Heidelberg, 2014 Predictability nationallicence Initialisation nationallicence Data assimilation nationallicence Southern Ocean nationallicence Sea ice nationallicence Zunz Violette Georges Lemaître Centre for Earth and Climate Research, Earth and Life Institute, Université catholique de Louvain, Louvain-la-Neuve, Belgium aut Goosse Hugues Georges Lemaître Centre for Earth and Climate Research, Earth and Life Institute, Université catholique de Louvain, Louvain-la-Neuve, Belgium aut Dubinkina Svetlana Centrum Wiskunde & Informatica, Amsterdam, The Netherlands aut Climate Dynamics Springer Berlin Heidelberg 44/7-8(2015-04-01), 2267-2286 0930-7575 44:7-8<2267 2015 44 382 https://doi.org/10.1007/s00382-014-2344-9 text/html Onlinezugriff via DOI BK010053 XK010053 XK010000 Metadata rights reserved Springer special CC-BY-NC licence nationallicence 1 research-article jats NATIONALLICENCE NATIONALLICENCE NL-springer NATIONALLICENCE 856 40 https://doi.org/10.1007/s00382-014-2344-9 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Zunz Violette Georges Lemaître Centre for Earth and Climate Research, Earth and Life Institute, Université catholique de Louvain, Louvain-la-Neuve, Belgium aut NATIONALLICENCE 700 1- Goosse Hugues Georges Lemaître Centre for Earth and Climate Research, Earth and Life Institute, Université catholique de Louvain, Louvain-la-Neuve, Belgium aut NATIONALLICENCE 700 1- Dubinkina Svetlana Centrum Wiskunde & Informatica, Amsterdam, The Netherlands aut NATIONALLICENCE 773 0- Climate Dynamics Springer Berlin Heidelberg 44/7-8(2015-04-01), 2267-2286 0930-7575 44:7-8<2267 2015 44 382