The modulation by the Atlantic multidecadal variability (AMV) of the dynamical climate response to a Pinatubo-like eruption is investigated for the boreal winter season based on a suite of large ensemble experiments using the CNRM-CM5 Coupled Global Circulation Model. The volcanic eruption induces a strong reduction and retraction of the Hadley cell during 2 years following the eruption and independently of the phase of the AMV. The mean extratropical westerly circulation simultaneously weakens throughout the entire atmospheric column, except at polar Northern latitudes where the zonal circulation is slightly strengthened. Yet, there are no significant changes in the modes of variability of the surface atmospheric circulation, such as the North Atlantic Oscillation (NAO), in the first and the second winters after the eruption. Significant modifications over the North Atlantic sector are only found during the third winter. Using clustering techniques, we decompose the atmospheric circulation into weather regimes and provide evidence for inhibition of the occurrence of negative NAO-type circulation in response to volcanic forcing. This forced signal is amplified in cold AMV conditions and is related to sea ice/atmosphere feedbacks in the Arctic and to tropical-extratropical teleconnections. Finally, we demonstrate that large ensembles of simulations are required to make volcanic fingerprints emerge from climate noise at mid-latitudes. Using small size ensemble could easily lead to misleading conclusions especially those related to the extratropical dynamics, and specifically the NAO.

This research was carried out within the pro- jects: (i) MORDICUS funded by the French Agence Nationale de la Recherche (ANR-13-SENV-0002-02); (ii) SPECS funded by the European Commission’s Seventh Framework Research Programme under the grant agreement 308378; (iii) VOLCADEC funded by the Spanish program MINECO/FEDER (ref. CGL2015-70177-R). We thank Javier Garcia-Serrano for its comments about the NAO precursors, Omar Bellprat for its suggestions concerning the statistical analysis and François Massonnet for its recommendations in terms of graphical presentation. CC is grateful to Marie-Pierre Moine, Laure Coquart and Isabelle Dast for technical help to run the model. Computer resources have been provided by Cerfacs. We thank the two anonymous referees for their useful comments and suggestions to improve this manuscript.

Peer Reviewed