Is the coupling of the jet stream strength in the Atlantic sector and NAO too weak in the circulation models?

Wintertime variability of both the strength of the jet stream and the North Atlantic Oscillation (NAO) index are known to be significantly correlated on the decadal scale and have positive trends since the 1960s which have been recently proposed to be connected to anthropogenic global warming (Blackport and Fyfe, 2022). At the same time there is a rich literature explaining both the observed variability and also the discrepancy with circulation models in which the variability is usually much smaller. The North Atlantic sector is known to be the nexus of the so called signal-to-noise paradox in climate modelling (Scaife and Smith, 2018) with models underestimating the interdecadal variability in both atmospheric circulation (NAO) and ocean temperatures (AMO/AMV) by an order of magnitude Smithe et al. 2020.

Scaife and Smith (2018) offer a selection of possible lacking processes causing this problem: (“lack of extratropical ocean–atmosphere coupling, weak eddy feedback in current resolution models, errors in remote teleconnections, or errors in parameterized processes such as atmospheric convection”. On the other hand, it is well known that the pattern of SST values on the North Atlantic and the position of the Gulf Stream affect the value (and sign) of wintertime NAO (Hermoso et al. 2024). The covariation of the jet stream strength and NAO on the decadal time scale seen in observation data suggests that this coupling may be one of the most important missing factors, however the phenomenon itself may be too weak in the models, which is one of the hypotheses to be tested.

This study tries is a first step in trying to find spatial patterns of differences in decadal-scale variability of atmospheric circulation at different altitudes in the Atlantic sector between (observation data assimilating) climate reanalyses and (non-assimilating) CMIP 5 & 6 runs The aim is to pinpoint where and how they differ. This presentation shows the preliminary results of such analysis.