We investigate the uncertainty (i.e., inter-model spread) in future projections of the northern winter climate, based on the forced response in the CMIP5 RCP8.5 scenario. The uncertainty in the forced response of sea level pressure (SLP) is large in the North Pacific, the North Atlantic, and the Arctic. A major part of these uncertainties (31%) is marked by a pattern with a center in the northeastern Pacific and a dipole over the northeastern Atlantic that we label as the Pacific–Atlantic SLP uncertainty pattern (PAΔSLP). To better understand the nature of PAΔSLP, the associated sea surface temperature (SST) and Arctic sea ice cover (SIC) perturbation patterns are prescribed in experiments with two atmospheric models (AGCMs): CAM4 and IFS.
The AGCM experiments indicate more robust SLP response over the North Pacific driven by the SST perturbation, which is associated with the tropical-midlatitude interaction and the Rossby wavetrain. The North Atlantic SLP response is better explained by the joint effect of SST and SIC perturbations, which is partly related to the Rossby wavetrain from the Pacific and the air–sea interactions over the North Atlantic. However, these responses shift westward relative to PAΔSLP, where in CAM4 it is related to the low-frequency transient eddy forcing. The magnitude of these responses in the two AGCMs is largely different. Thus, constraining only the SST and SIC projections might not alone help constrain future climate projections. We should investigate the role of other factors in these uncertainties, such as the atmosphere–SST–SIC coupled dynamics.