Dr. Jerry Meehl
Abstract: Decadal climate prediction presumes there are relevant decadal timescale processes and mechanisms that, if initialized properly, provide predictive skill beyond the first year or two. Candidate mechanisms involve Pacific Decadal Variability (PDV, e.g. the Interdecadal Pacific Oscillation, IPO), and Atlantic Multidecadal Variability (AMV, e.g the Atlantic Multidecadal Oscillation, AMO). If one is driving the other, then a skillful prediction of decadal SST variability in one basin would result in skillful predictions of SST in the other basin, with resulting global-scale teleconnections that would then simplify the decadal climate prediction problem. Here we use pacemaker model configurations with specified observed patterns of SSTs for AMV and PDV, with the rest of the model fully coupled, to show that there tends to be a weak opposite-sign SST response in the tropical Pacific when SSTs are specified in the Atlantic. Conversely, for the PDV experiment, there is a weak same-sign SST response in the tropical Atlantic. Net surface heat flux in the Atlantic and ocean dynamics in the Pacific play contrasting roles in the ocean response in the respective basins. We propose a new paradigm for decadal timescale variability such that processes in the Pacific and Atlantic are mutually interactive through the atmospheric Walker Circulation. This result implies that processes and mechanisms in both basins, and their interactions, must be simulated to produce credible decadal climate predictions.