LAUSR

The response of the European climate to the Atlantic multidecadal variability (AMV) remains difficult to isolate in observations because of the presence of strong internal variability and anthropogenically forced signals. Using model sensitivity experiments proposed within the CMIP6/Decadal Climate Prediction Project Component C (DCPP-C) framework, the wintertime AMV–Europe teleconnection is here investigated in large ensembles of pacemaker-type simulations conducted with the CNRM-CM5 global circulation model. To evaluate the sensitivity of the model response to the AMV amplitude, twin experiments with the AMV forcing pattern multiplied by 2 and 3 (2xAMV and 3xAMV, respectively) are performed in complement to the reference ensemble (1xAMV). Based on a flow analog method, we show that the AMV-forced atmospheric circulation tends to cool down the European continent, whereas the residual signal, mostly including thermodynamical processes, contributes to warming. In 1xAMV, both terms cancel each other, explaining the overall weak AMV-forced atmospheric signal. In 2xAMV and 3xAMV, the thermodynamical contribution overcomes the dynamical cooling and is responsible for milder and wetter conditions found at large scale over Europe. The thermodynamical term includes the advection of warmer and more humid oceanic air penetrating inland and the modification of surface radiative fluxes linked to both altered cloudiness and snow-cover reduction acting as a positive feedback with the AMV amplitude. The dynamical anomalous circulation combines 1) a remote response to enhanced diabatic heating acting as a Rossby wave source in the western tropical Atlantic and 2) a local response associated with warmer SST over the subpolar gyre favoring an anomalous high. The extratropical influence is reinforced by polar amplification due to sea ice melting in all the subarctic seas. The weight between the tropical–extratropical processes and associated feedbacks is speculated to partly explain the nonlinear sensibility of the response to the AMV forcing amplitude, challenging thus the use of the so-called pattern-scaling technique to evaluate teleconnectivity and related impacts associated with decadal variability.