Model-Dependent Atlantic Multidecadal Variability Modulations on North Pacific Sea Surface Temperature Variability and Decadal Prediction

Previous studies suggested that Atlantic Multidecadal Variability (AMV) modulations on pan-Pacific sea surface temperature (SST) variability and prediction are model-dependent. These results were mainly based on SST forcing experiments in which AMV-related Atlantic SST anomalies were prescribed. However, the AMV itself is also model-dependent, but its influences on the Pacific remain unclear. Here, we use multi-model fully coupled experiments from the Coupled Model Intercomparison Project Phase 6 (CMIP6), along with observations, to study the model-dependent AMV trans-basin effects. We found that AMV strength is a key factor: Stronger (Weaker) model AMV than observations overestimates (underestimates) SST response and decadal prediction skills, mainly in the North Pacific. The reason is that stronger positive phased AMV, for example, leads to higher sea level pressure anomalies over the North Pacific, which lifts sea surface height and deepens thermocline to warm SST. Our study highlights the necessity to improve simulations of AMV strength. Pacific sea surface temperature (SST) decadal variability and prediction are important for economy and environment. The Atlantic Multidecadal Variability (AMV) was thought to significantly influence Pacific SST variability and prediction. However, the results of previous studies were model-dependent. In this paper, we use fully coupled experiments of the Coupled Model Intercomparison Project Phase 6 (CMIP6) models to study what contributes to the model dependency. We found that the strength of the AMV is a key factor influencing AMV trans-basin modulations on SST variability and decadal prediction, mainly in the North Pacific Ocean. The related mechanisms are discussed. Our paper sheds light on the way to improve North Pacific SST decadal prediction. Climate models show a strong inter-model spread of Atlantic Multidecadal Variability (AMV) strengthModels with stronger (weaker) AMV than observations overestimate (underestimate) North Pacific sea surface temperature (SST) response and decadal predictionThe North Pacific SST response is primarily forced by AMV-induced wind stress curl and ocean dynamics