Abstract
Predictability of atmospheric variability is known to be limited owing to significant uncertainty that arises from intrinsic variability generated independently of external forcing and/or boundary conditions. Observed atmospheric variability is therefore regarded as just a single realization among different dynamical states that could occur. In contrast, subject to wind, thermal and fresh-water forcing at the surface, the ocean circulation has been considered to be rather deterministic under the prescribed atmospheric forcing, and it still remains unknown how uncertain the upper-ocean circulation variability is. This study evaluates how much uncertainty the oceanic interannual variability can potentially have, through multiple simulations with an eddy-resolving ocean general circulation model driven by the observed interannually-varying atmospheric forcing under slightly different conditions. These ensemble “hindcast” experiments have revealed substantial uncertainty due to intrinsic variability in the extratropical ocean circulation that limits potential predictability of its interannual variability, especially along the strong western boundary currents (WBCs) in mid-latitudes, including the Kuroshio and its eastward extention. The intrinsic variability also greatly limits potential predictability of meso-scale oceanic eddy activity. These findings suggest that multi-member ensemble simulations are essential for understanding and predicting variability in the WBCs, which are important for weather and climate variability and marine ecosystems.
Highlights
As the average over the three ensemble members, time-mean values and interannual variance of SSH are overall reproduced well in the model with respect to their geographical distributions and amplitudes (Supplementary Fig. S4)
We use the Modular Ocean Model 3 OGCM27 with substantial modifications added for its optimal performance on the vector-parallel architecture of Japan’s Earth Simulator
A particular flow field simulated in the 16th year of that integration was used as the initial condition for our ensemble hindcast integration that is forced by six-hourly atmospheric fields taken from the Japanese 25-year Reanalysis[32] for 1979 to 2012
Summary
As the average over the three ensemble members, time-mean values and interannual variance of SSH are overall reproduced well in the model with respect to their geographical distributions and amplitudes (Supplementary Fig. S4). Some discrepancies are noticeable between the simulated and observed fields. In the simulated time-mean field, for example, detailed structures of the subtropical frontal zone in the western North Pacific (20–25 °N) are missing, and SSH gradients under the Intertropical Convergence Zone (around 5–10 °N) are somewhat overestimated. The SSH variance is underestimated east of the Mindanao Island and in the eastern North Pacific. Despite these discrepancies, the overall model performance seems excellen
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
