Atlantic and Pacific Tropics Connected by Mutually Interactive Decadal-Timescale Processes
Decadal climate prediction presumes there are decadal timescale processes and mechanisms that, if initialized properly in models, potentially provide predictive skill more than one or two years into the future. It is possible that the tropical Atlantic could drive the tropical Pacific or vice versa, and predicting decadal variability in one could provide prediction skill in the other. We analyze observations and model simulations where sea surface temperatures are specified in one region and the rest of the coupled model responds to isolate the effects of one ocean basin on the other.
There tends to be a weak opposite-sign SST response in the tropical Pacific when observed SSTs are specified in the Atlantic, while there is a weak same-sign SST response in the tropical Atlantic when observed SSTs are specified in the tropical Pacific. Therefore, one basin does not drive the other, but processes in the Pacific and Atlantic are sequentially and mutually interactive through the atmospheric Walker Circulation along with contributions from midlatitude teleconnections for the Atlantic response to the Pacific
The implications for decadal climate prediction are that processes and mechanisms in both the Atlantic and Pacific must be simulated and predicted since each basin depends on the other for prediction skill on decadal timescales.