Understanding intrinsic ocean variability by suppressing regional stochastic variability

Intrinsic ocean variability is essential for climate prediction because it is less sensitive to stochastic process, but it is very difficult to be identified due to internal climate variability. Here we use regional interactive ensemble applied on ocean-atmosphere interface (RIE-OA) to suppress atmosphere stochastic variability and to reveal intrinsic variability as well as to understand climate dynamic across multiple timescales. Five atmosphere general circulation models (AGCM) are coupled to an ocean general circulation model (OGCM) over the North Atlantic basin (20^oN to Denmark Strait and Greenland-Scotland ridge). The OGCM interacts with fluxes from a selected AGCM globally except over the North Atlantic basin where the OGCM interacts with the ensemble averaged fluxes from the five AGCMs. The five AGCMs, on the other hand, feel the same ocean states. Hence, the atmosphere stochastic variability impacting the ocean is one-fifth weaker than stand-alone configuration (control case). This leads to reduction of the local climate variability, such as Atlantic Multidecadal Variability, but should not reduce intrinsic variability. Comparing control cases and RIE-OA case, we found the intrinsic ocean variability, a narrow-banded low-frequency (about 8 to 20 years) signal over the North Atlantic Subtropical Gyre, is not influenced by the weakened stochastic variability. More details will be discussed.